Diplomarbeit · Ich erkläre weiters Eides statt, dass ich meine Diplomarbeit nach den anerkannten Grundsätzen für wissenschaftliche Abhandlungen selbstständig ausgeführt habe

Diplomarbeit

QUANTIFICATION OF TOTAL CYANIDE CONTENT

IN KERNELS OF STONE FRUITS

ausgeführt zum Zwecke der Erlangung des akademischen Grades eines

Diplom-Ingenieursunter der Leitung von

Ao.Univ.Prof. Dipl.-Ing. Dr.techn. Ingrid Steiner(E330 Institut für Managementwissenschaften, Bereich: Betriebstechnik und Systemplanung)

eingereicht an der Technischen Universität Wien

Fakultät für Technische Chemie

von

Anatol Maurice Desser9625123

Reizenpfenninggasse 10/6

1160 Wien

Wien, im März 2015 _________________________

Anatol Maurice Desser

QUANTIFICATION OF TOTAL CYANIDE CONTENT IN KERNELS OF STONE FRUITS

Ich habe zur Kenntnis genommen, dass ich zur Drucklegung meiner Arbeit unter derBezeichnung

Diplomarbeitnur mit Bewilligung der Prüfungskommission berechtigt bin.

Ich erkläre weiters Eides statt, dass ich meine Diplomarbeit nach den anerkannten Grundsätzenfür wissenschaftliche Abhandlungen selbstständig ausgeführt habe und alle verwendetenHilfsmittel, insbesondere die zugrunde gelegte Literatur, genannt habe.

Weiters erkläre ich, dass ich dieses Diplomarbeitsthema bisher weder im In- noch Ausland(einer Beurteilerin/einen Beurteiler zur Begutachtung) in irgendeiner Form als Prüfungsarbeitvorgelegt habe und dass diese Arbeit mit der vom Begutachter beurteilten Arbeit übereinstimmt.

Wien, im März 2015 _________________________Anatol Maurice Desser


Danksagung

An dieser Stelle möchte ich mich bei meinen Eltern Dr. Lucia und Dr. Hans Desser bedanken, diemir durch ihre finanzielle und moralische Unterstützung - besonders in schwierigen Situationen- mein Studium ermöglicht haben.

Für ihre hilfreiche Unterstützung bei meiner Arbeit möchte ich mich auch den Mitarbeitern derArbeitsgruppe Lebensmittelchemie für die gute Zusammenarbeit, ganz besonders meinenKollegen Bernd Trimmel und DI Kathrin Scharnhorst, meiner Betreuerin Prof. Dr. Ingrid Steinerund Herbert Hafner bedanken.

Mein Dank gilt ebenfalls Dr. Gerd Mauschitz für seine Hilfe bei der Laserdiffraktometrie.


Abstract

Many plants from the genus prunus deliver fruits which are popular food or pre productin many parts of the world. The kernels of those fruits vastly remain unprocessed andare disposed of, because of the high content of cyanogenic glycosides which refers to upto 2500 mg HCN per kg plant material. Therefore it is crucial to be aware of the totalcyanide content and to apply appropriate methods of decontamination to allowindustrial use.

In this work the method of determination of total cyanide with picric acid was adaptedfor analysis of stone fruit kernels. The content of total cyanide was determinedphotometrically according to the reaction of HCN with picric acid to isopurpuric acid ona test strip and expressed in mg HCN per kg food sample.

In the different stone fruit kernels (apricot, sort cherry and bitter almond) the contentof total cyanide was determined to be in a range of 800 to 2400 mg/kg. In stone fruitkernels from sweet cultivars (sweet almond) no cyanide was determined. In productsmade from stone fruit kernels (roasted kernels, persipan) cyanide was determined to bein a range of 60-173 mg per kg food sample. In other products made from stone fruitkernels (marzipan, marzipan filling, chocolate) no cyanide was detected.

Several methods of decontamination were tested: continuous flow and batch by water,batch by ethanol, heat and fermentation with β-glucosidase. It was shown that satisfyingdecontamination can be achieved simply with cold water. The HCN content of apricotkernels was reduced by 80% in a 24h batch process using cold water.

The results of this study indicate that stone fruit kernels and kernel products can beused in the food industry after an appropriate decontamination step. Furthermore it wasshown that the method of determining the HCN content with picric acid on a test stripcan be used for rapid analysis.

4


Contents

1. Introduction and objectives ................................................................................................................................... 11

1.1 Introduction ........................................................................................................................................................... 11

1.2 Objectives ................................................................................................................................................................ 11

2. Principles ...................................................................................................................................................................... 12

2.1 Cyanogenic glycosides and cyanogenesis ................................................................................................... 12

2.1.1 General .................................................................................................................................................................. 12

2.1.2 Biosynthesis ....................................................................................................................................................... 13

Biosynthesis of dhurrin and amygdalin ............................................................................................................. 15

2.1.3 Occurrence ......................................................................................................................................................... 16

2.1.4 β-Glucosidases .................................................................................................................................................. 17

2.1.5 Toxicity ................................................................................................................................................................. 17

2.5.1.1 Absorption, distribution and excretion ............................................................................................... 17

2.5.1.2 Toxic Effects .................................................................................................................................................. 18

2.1.5.3 Mechanism ..................................................................................................................................................... 18

2.1.5.4 Detoxification ................................................................................................................................................ 18

2.1.5.5 Symptoms of poisoning ............................................................................................................................. 19

2.1.5.6 Treatment ...................................................................................................................................................... 19

2.1.6 Physical and chemical properties of stone fruit kernels .................................................................. 19

2.1.6.1 Physical and chemical properties of apricot kernels ..................................................................... 19

2.1.6.2 Chemical composition of sour cherry kernel oil ............................................................................. 21

2.2 Determination of cyanogenic potential in stone fruit kernels and their products. ................... 22

2.2.1 Norm method: determination of hydrocyanic acid in marzipan, persipan and theirprestages by distillation and titration. ................................................................................................................ 22

2.2.2 Norm Method: Animal feeding stuffs — Determination of Hydrocyanic acid by HPLC ...... 22

Range of application ................................................................................................................................................... 22

2.2.3 Determination of cyanide content by chemical or enzymatic hydrolysis of the cyanogenicglycosides (CG) and quantification of cyanide. ................................................................................................ 22

2.2.4 Determination of amygdalin by chromatographic methods .......................................................... 23

2.2.4.1 Determination of amygdalin by high pressure liquid chromatography HPLC ................... 23

2.2.4.2 Determination of amygdalin by gas chromatography GC/MS ..................................................... 23

2.2.4.3 Determination of amygdalin by micellar electrokinetic chromatography (MEKC) ............ 23

2.2.5 Determination of amygdalin by enzyme-immunoassay ................................................................... 24

5


2.3 Decontamination methods ............................................................................................................................... 24

2.3.1 Decontamination by water ........................................................................................................................... 24

2.3.2 Decontamination with ethanol ................................................................................................................... 25

2.3.3 Decontamination by fermentation with emulsine (autodecontamination) ............................. 25

2.4 Processing options of stone-fruit kernels containing cyanogenic compounds ........................... 26

2.4.1 Overview .............................................................................................................................................................. 26

2.4.1. Separation of the apricot the kernels from the pits ........................................................................... 27

2.4.2 Kernel processing ............................................................................................................................................ 30

2.4.2.1 Production of edible apricot oil ............................................................................................................... 30

2.4.2.2 Production of natural benzaldehyde from stone fruit kernels ................................................... 30

2.4.2.2.1 Principles ...................................................................................................................................................... 30

2.4.2.2.2 Production of natural benzaldehyde ................................................................................................. 31

2.4.2.2.2.1 Recovery of wastewater containing benzaldehyde by simultaneous distillation-extraction (SDE) ........................................................................................................................................................... 31

2.4.2.2.2.2 Base-catalysed production from natural cinnamaldehyde ................................................. 31

2.4.2.2.2.3 Extraction by microwave .................................................................................................................. 31

2.4.2.2.2.4 Quantification of benzaldehyde in the product. ........................................................................ 31

2.4.2.3 Production of persipan ............................................................................................................................... 31

2.4.2.4 Antioxidant properties of roasted apricot kernel flours ............................................................... 32

2.4.3 Processing kernels from other stone fruits ............................................................................................ 32

3. Materials ........................................................................................................................................................................ 33

3.1 Sample Materials .................................................................................................................................................. 33

3.2 Laboratory devices .............................................................................................................................................. 34

3.3 Chemicals ................................................................................................................................................................ 34

4. Methods ......................................................................................................................................................................... 35

4.1 Picrate method for determination of total cyanide in prunus kernels ........................................... 35

4.1.1 Principle ............................................................................................................................................................... 35

4.1.2 Sample Preparation and preparation of picrate paper test strips ................................................ 35

4.1.2.1 Preparation of picrate paper test strips ............................................................................................... 35

4.1.2.2 Preparation of homogeneous food samples like marzipan or persipan ................................. 35

4.1.2.3 Preparation of apricot kernels ................................................................................................................. 36

4.1.2.4 Preparation of sour cherry kernels ....................................................................................................... 36

4.1.3 Quantification of the total cyanide content in the prepared sample ............................................ 36

4.1.4 Method validation ............................................................................................................................................ 37

4.1.4.1 Working range ................................................................................................................................................ 38

4.1.4.2 Linearity ........................................................................................................................................................... 38

6


4.1.4.3 Limit of detection .......................................................................................................................................... 38

4.1.4.4 Limit of quantification ............................................................................................................................... 38

4.1.4.5 Precision .......................................................................................................................................................... 38

4.1.4.6 Selectivity ........................................................................................................................................................ 39

4.2 Decontamination .................................................................................................................................................. 40

4.2.1 Decontamination using water .................................................................................................................... 40

4.2.1.1 Continuous flow decontamination ......................................................................................................... 40

4.2.1.2 Batch decontamination .............................................................................................................................. 40

4.2.1 Decontamination using ethanole ............................................................................................................... 40

4.2.3 Decontamination by heat treatment ......................................................................................................... 41

5. Results ............................................................................................................................................................................ 42

5.1 Preliminary tests and observations .............................................................................................................. 42

5.1.1 Apricot .................................................................................................................................................................. 42

5.1.2 Sour Cherry ......................................................................................................................................................... 42

5.2 Total HCN content in stone fruit kernels and foods containing these ............................................. 43

5.3 Effects of debittering .......................................................................................................................................... 45

5.3.1 Effects of debittering by leaching in water ............................................................................................. 45

5.3.1.1 Continuous flow debittering .................................................................................................................... 45

5.3.1.2 Batch debittering .......................................................................................................................................... 45

5.3.2 Effects of debittering by leachking in ethanol ....................................................................................... 46

5.3.3 Effects of debittering by heat treatment ................................................................................................. 46

5.4 Particle size distribution .................................................................................................................................. 46

6. Discussion ..................................................................................................................................................................... 50

6.1 Method ..................................................................................................................................................................... 50

6.2 Sustainabilty of stone fruit kernels ............................................................................................................... 50

6.3 Toxicity of stone fruit kernel products ........................................................................................................ 50

7. Abbreviations .............................................................................................................................................................. 52

References ......................................................................................................................................................................... 54

7


1. Introduction and objectives

1.1 IntroductionDue to the natural limitation of non-renewable resource it is generally sensible toachieve new and more effective possibilities of using renewable primary products if thisis technically possible and economically reasonable. Stone fruit kernels are rich in edibleoil of high quality, proteins and fibres and therefore an example of a vast unused sourcefor renewable primary products. They can be used in production of edible oil, cosmetics,drugs and scents, whereas the pits can be used for fuel and heat production (1) (2).

Since the use of stone fruit kernels is connected with some technical problems, such ascrushing the stones properly without damaging the kernel, separation of the kernel fromthe pit (1) and detoxification of the amygdalin-containing kernel, most stones producedin the EU are composted or used as fuel for heat production (2), without separating andprocessing the kernel to valuable secondary products.The indirect toxicity of amygdalin which is decomposed by the enzyme ß-glucosiadase tobenzaldehyde, glucose and prussic acid represents a potential danger for the consumer. The Austrian food law allows 50 mg/kg in nougat, marzipan, alternative products andsimilar products, 5 mg HCN/kg in stone fruit preserves and 35 mg/kg in alcoholic drinks(3).1 mg hydrocyanic acid corresponds to approximately 17 mg amygdalin. Because thedecomposition product natural benzaldehyde (bitter almond oil) is an importantflavouring substance for various aromas it is important to learn more about thedecomposition process in order to optimize the detoxification of stone fruit kernels fromhydrocyanic acid and the utilisation of benzaldehyde.

1.2 ObjectivesThe purpose of this work is to discuss and evaluate different methods for thedetermination of amygdalin and the total cyanide content in different matrices and todevelop a rapid test for quantification of the total cyanide content in stone fruit kernelsand their products. A further purpose is to discuss the options, challenges and dangers of utilizing stonefruit kernels, including a discussion on methods of detoxification.

8


2. Principles

2.1 Cyanogenic glycosides and cyanogenesis

2.1.1 GeneralHydrogen cyanide (HCN) formation in plants has been known since ancient ages. Inancient Egypt, traitorous priests in Memphis and Thebes were poisoned to death withkernels of peaches (4) (5). In 1802, the pharmacist Bohm detected liberated HCN bydistillation of bitter almonds (4) (6). In 1830, Robiquet and Boutron-Chalard discoveredthe structure of the cyanogenic compound in bitter almonds (4) (6). It was namedamygdalin because the compound was isolated from Prunus amygdalus (synonym forPrunus dulcis or almond). Amygdalin has been found in many different seeds of othermembers of the Rosaceae like in apples (Malus spp.), peaches (Prunus persica), apricots(Prunus armeniaca), black cherries (Prunus serotina), and plums (Prunus spp.) (7) (8)(9) (10) (11) (12) (13) (14). The diglucoside amygdalin was the first cyanogenicglucosides which was isolated. Cyanogenic glucosides are present in at least 2650species from more than 550 genera, and 130 families (15) (16), including manyimportant crop plants (17) (18). After destruction of plant tissue containing cyanogenicglucosides, these are hydrolyzed by beta-glucosidases to glucose, an aldehyde or ketone,and HCN. This two-component system, of which each of the separate components ischemically inert, provides plants with an immediate chemical defence against attackingherbivores and pathogens (19) (20) (21) (22) (23) (24) (25) (26). Besides their possibledefence function, accumulation of cyanogenic glucosides in certain angiosperm seedsmay act as a storage deposit of reduced nitrogen and sugar for the developing seedlings(27) (28) (29) (11). Cyanogenic glycosides are the most common known forms of boundhydrocyanic acid. Besides cyanogenic glycosides, cyanolipids are O-beta-glycosidedalpha-hydroxynitriles (cyanhydrines), which occur mainly in plants and in some insects(30).The most published compounds are glycosides of α-hydroxynitriles or cyanohydrinesand a few cyanogenic lipids. The glyocosides are usually accompanied by a β-glycosidaseenzyme which can produce the corresponding cyanohydrine (aglycon) and sugar. Asecond enzyme, the hydroxynitrile lyase catalyzes the dissociation of the cyanohydrineto a carbonyl compound and hydrogen cyanide (16) (Figure 2.1). Usually the substrateand the enzymes are stored in different compartments within the plant cell, thereforecyanide release does not occur without damage of the plant (31) (32).

9


Figure 2.1: β-glycosidase produces the corresponding cyanohydrine (aglycon) andsugar. Hydroxynitrilelyase catalyses the dissociation of the cyanohydrine to a carbonylcompound and hydrogen cyanide (16).

The production of cyanide in many plants is extremely variable. Nevertheless, thedistribution of cyanogenesis by cyanogenic glycosides is of systematic interest becausecertain structural types of cyanogenic glycoside are associated with specific groups ofplants (15) (16)(33).Cyanogenic glycosides are classified by their biogenetic precursors. These are theproteinogenic amino acids L-valine, L-isoleucine, L-leucine, L-phenylalanine and L-tyrosine. Furthermore the non-proteinogenic amino acid (2-cyclopentyl)-glycin works asprecursor (34) for some cyanogenic glycosides such as deidaclin (35) and probablyniacin as precursor of acalyphin (30) (36).

10


Cyanogenic glycosides are found in fern, gymnosperms and angiosperms. In ferns onlycompounds of phenylalanine-derived types were detected, in gymnosperms onlytaxiphyllin which is a cyanogenic glycoside derived of tyrosine. Angiosperms show a vastspread of cyanogenic glycosides. In monocotyledons the tyrosine-derived types aredominant whereas in dicotyledons almost all known cyanogenic glycosides are found.Currently more than 62 cyanogenic compounds are known (15)(35).Hydrocyanic acid can be bound in form of cyanolipids by the family of sapindaceae andhippocastanaceae. In cyanolipids α-hydroxynitrile is bound as ester with one or two,mostly C18:1 or C18:2 fatty acids (35).In almost every plant, cyanogenic or non-cyanogenic, it is possible to detect smallamounts of free hydrogenic acid. This free HCN is mainly a by-product of ethyleneproduction and is not based on cyanogenesis and therefore should not to be perceived ascyanogenic processes (35).Respectively to the biological and ecological meaning of cyanogenic glycosides in plantsthe function as defence mechanism against herbivore is to be noted(37). Highconcentrations of cyanogenic glycosides are found especially in organs of reproduction(flower, fruit, and seed) and in young leaves, so those parts of plants that need mostprotection due to their function and their stage of development. These facts attest to thethesis of a defensive role of cyanogenesis (30).Cyanogenesis in agricultural plants such as cassava (Manihot esculenta, Euphorbiaceae)is of toxicological relevance for human nutrition (38) (39). After intake of cyanogenicglycosides, freed hydrocyanic acid acts neurotoxically. By degenerating periphericalneurons, damages of the nervus opticus and dyskinesia of extremities can occur (35)(30) (40). Thiocyanate formed from cyanide inhibits the production of the thyroid hormonethyroxine and can lead to formation of goiters. Pharmaceutical products that containcyanonogenic glycosides (e.g. Laetrile ®) are not permitted in the European Union (35).

2.1.2 BiosynthesisCyanogenic glycosides are secondary plant compounds that occur vastly in many plants.They release HCN which can make the plant poisonous if eaten. The enzymesresponsible for production of the HCN have been known for long. The aglycone portion of cyanogenic compounds is derived from amino acids. Mostcyanogenic compounds come from a relatively small number of precursors: L-tyrosine,L-phenylalanine, valine, isoleucine, and L-leucine. Acalyphin appears to be derived fromnicotinic acid. One series of cyanogens has aglycones that contain a cyclopentenoid ringand are derived from the non-protein amino acid, (2-cyclopentenyl)glycine (16) (41)(30).By double labelling experiments it has been demonstrated that the C-N bond is notbroken during the biosynthetic process. All intermediates in the pathway containnitrogen. (16)(42). Major intermediates include a hydroxyl amino acid, aldoxime,nitrile, and a α-hydroxynitrile (16) (43) (44). These intermediates are observed only atextremely low levels in the plants that contain them and that the proposed intermediateswere incorporated at very different and inconsistent rates. These differences in rateswere inconsistent with the order of incorporation of the different intermediates. A

11


channelled process has been proposed in which the compounds are not liberated fromthe enzyme surface (16)(31). Channelling provides for the rapid and efficient flow ofcarbon and nitrogen atoms and at the same time perhaps protects labile intermediatesfrom wasteful side reactions (16)(31).β-Glucosyltranferases have been isolated and purified from a few cyanogenic species.They have pH-optima from 6.5-9.0 and usually lack a requirement for metal ions orcofactors. Although they have an absolute specifity for UDP-glucose, they are less specifictoward cyanohydrins (16)(45).

Precursor Cyanogenic Glycosides (examples)

L-phenylalanine Diglycosides:(R)-amygdalin, (R)-sucumin, (S)-epilucumin, (R)-vicianin, grayanin (46)Monoglucosides: (R)-prunasin, prunasin-6-malonate (47)

L-valine and L-isoleucine linamarin, (R)-lotaustarlin, (S)-epilotaustralin, linustatin

L-leucine (S)-proacacipetalin, (R)-epiproacacipetalin, cyanolipids

L-tyrosine (S)-dhurrin, (R)-taxiphyllin, (S)-heterodendrin

(2-cyclopentenyl)glycin deidaclin, tertraphyllin A and B, volkenin,taraktophyllin

Niacin scalyphin (probably)

Table 2.1.: Some cyanogenic compounds and their precursors (16)

Biosynthesis of dhurrin and amygdalinStudies in almonds have shown that prunasin is transformed into amygdalin during fruitripening (9). Prunasin is present in roots, leaves, and kernels of sweet and bittervarieties (14). By tracer experiments it has been demonstrated that prunasin issynthesized from phenylalanine (48). Sorghum (Sorghum bicolor) contains the tyrosine-derived cyanogenic glucoside dhurrin (49) (45). Dhurrin biosynthesis is catalyzed by themultifunctional and membrane-bound enzymes cytochrome P450 (Cyt P450), CYP79A1and CYP71E1 (50) (51) (52) (53). CYP79A1 catalyzes the conversion of tyrosine into Z-p-hydroxyphenylacetaldoxime (51) and CYP71E1 catalyzes the conversion of the Z-p-hydroxyphenylacetaldoxime into p-hydroxymandelonitrile (54) (52). Conversion of thelabile cyanohydrin into dhurrin is catalyzed by the soluble enzyme UDP-Glc (UDPG)-glucosyltransferase UGT85B1 (55) (56) (See Fig2.2). The entire pathway for dhurrinsynthesis has been transferred from sorghum to arabidopsis (Arabidopsis thaliana)using genetic engineering (57) (58). Prunasin biosynthesis is assumed to follow thesame biosynthetic scheme as dhurrin biosynthesis but no enzymes or genes involved

12


have been identified yet (10). The monoglycoside prunasin is converted into thediglucoside amygdalin by an additional UDPG-glucosyltransferase (Fig. 2.1). In contrastto the biosynthetic enzymes in almond, the enzymes and genes involved in amygdalindegradation have been identified and characterized. The first step in amygdalindegradation is catalysed by the β-glucosidase amygdalin hydrolase (EC 3.2.1.117) andresults in the formation of prunasin and glucose. Prunasin is subsequently hydrolyzed byprunasin hydrolase (EC 3.2.1.21), another β-glucosidase to form mandelonitrile andglucose. Mandelonitrile then is converted into benzaldehyde and HCN by mandelonitrilelyase (EC 4.1.2.10; (59). This conversion may also proceed non-enzymatically at neutralor alkaline pH. HCN inhibits cell respiration and is detoxified by β-cyano-Ala synthase(CASase), which converts HCN and cysteine into b-cyano-alanine (60). β-Cyano-alanineis converted into asparagine and aspartic acid by nitrilases (4)(61) (62) (63) (64) (65).

Figure 2.2: The metabolic pathways for synthesis and catabolism of the cyanogenic glucosidesprunasin and amygdalin in almonds. Biosynthetic enzymes (black lines) are: CYP79 and CYP71,Cyt P450 monooxygenases; GT1, UDPG-mandelonitrile glucosyltransferase; and GT2, UDPG-prunasin glucosyltransferase. Catabolic enzymes (dashed lines) are: AH, amygdalin hydrolase;PH, prunasin hydrolase; MDL1, mandelonitrile lyase; and ADGH*, amygdalin diglucosidase(putative). (4)

2.1.3 Occurrence Cyanogenic compounds are found in all major plant groups. The distribution ofcyanogenic compounds has been reviewed (33) (66) (67) (68). In families Araceae,Asteraceae, Euphorbiaceae, Fabaceae (69), Flacourtiaceae, Malesherbiaceae,Papaveraceae, Passifloraceae, Poaceae, Proteaceae, Ranunculaceae, Rosaceae,Sapindaceae and Turnerceae cyanogenesis is most common (16).There is much variation from plant to plant in concentration of cyanogenic glycosidesand β-glucosidases necessary to hydrolyze them. In a number of taxa only 1-3% of theindividuals will have positive cyanogenic test results, in others all individuals arecyanogenic (16)(70) .

13


In some genera, almost all tested species are cyanogenic. Only a few species of thegenera Prunus (Rosaceae) or Passiflora (Passifloraceae) are not cyanogenic. In someother families and genera, only a single species is cyanogenic (15) (16). Cyanogenic glycosides derived from L-phenylalanine, such as amygdalin, are foundprimarily in Rosidae and Asteridae. The compound best known is amygdalin, which iswidespread in seeds of members of the Rosaceae such as apples, peaches, cherries andapricots (16). Although it has been thought previously, that the Rosaceae are relatively homogeneouswith regard to cyanogenesis and that the major compounds are derived fromphenylalanine, this appears now to be true only for the subfamilies Maloideae andAmydaloideae. Other members of the family contain compounds derived from otherprecursors such as the p-hydroxybenzoate and p-hydroxycinnamate of (S)-cardiospermin and dhurrin (16).

2.1.4 β-GlucosidasesIn most plants cyanogenic glycosides are accompanied by a corresponding β-glycosidase. One such enzyme system, emulsine from almond seeds, has fairly broadspecificity and has been studied widely. Most of these enzymes have more specificsubstrate requirements. The specificity of enzymes from plants containingcyclopentenoid cyanogenic glycosides has been reviewed (71).Also the activity of β-glycosidase can differ from plants containing diglycosides. E.g. theenzymes from Davallia trichomaoides and Vicia angustifolia hydrolyze (R)-viciantin and(R)-amygdalin at the aglycone-disaccharide bond, and the enzymes of Prunin serotinaand Linum usitatissimum involve a stepwise removal of the sugars (72).In some instances, when an unnatural enzyme is added to a substrate before the naturalenzyme, no reaction is observed. The nature of these effects is dependent onconcentrations of enzymes and substrates as well as the order of addition. Interactionsof this type for cyclopentenoid cyanogenic glycosides and the corresponding enzymesystem have been reviewed (71).

2.1.5 ToxicityCyanide poisoning occurs when a living organism is exposed to a compound thatproduces cyanide ions. The toxicity of cyanogenic glycosides is caused by the release offree hydrogen cyanide when they are decomposed. The cyanide ion halts cellularrespiration by inhibiting the enzyme cytochrome c oxidase (EC 1.9.3.1) in themitochondria. The Scientific Panel on Food Additives, Flavourings, Processing Aids and Materials inContact with Food (AFC) noted that cyanogenic glycosides present in plants, as sourcesof hydrocyanic acid, are relatively nontoxic until HCN is released. This can occur as aresult of enzymatic hydrolysis by β-glucosidases following maceration of plant tissue, orby the gut microflora. Depending on the specific glycosides the hydrolysis products canbe, besides sugar moieties and HCN, benzaldehyde (for amygdalin, prunasin,sambunigrin,), phydroxybenzaldehyde (for dhurrin) and acetone (for linamarin). The

14


potential toxicity of a cyanogenic plant depends primarily on its capacity to produce HCN(73). In summary the Panel noted further that:1. Following oral administration, HCN is readily absorbed and rapidly distributed in thebody via the blood.2. HCN absorbed from the gut is metabolically converted to the less toxic thiocyanate.Other detoxification pathways include combination with vitamin B12 or some sulphur-containing amino acids. Acute toxicity results when the rate of absorption of HCN is suchthat the metabolic detoxification capacity of the body is exceeded.3. Cases of human intoxication and chronic neurological effects have occurred from theingestion of processed plants. This is particularly apparent when processing practiceschange as a result of trading practices or uncertain food supply.4. The cyanide ion inhibits enzymes associated with cellular oxidation and causes deaththrough energy deprivation. The symptoms, which occur within a few minutes, mayinclude constriction of the throat, nausea, vomiting, giddiness, headache, palpitations,hyperpnoea then dyspnoea, bradycardia, unconsciousness and violent convulsions,followed by death.5. The occurrence of intoxication symptoms depends upon the rapidity of the increase inHCN concentration in the tissues. In the case of cyanogenic glycosides, the route ofexposure, the nature of the cyanogenic compound, the dose, and the ability of theorganism to detoxify cyanide determine the symptoms.6. The available toxicity data show that cyanogenic glycosides from certain plants couldproduce acute toxic effects. Thus fatalities have occurred from e.g. consumption of stonefruit kernels.7. The chronic uptake of HCN, in sub-acutely toxic doses, may be involved in thepathogenesis of certain conditions including disturbance of thyroid function andneuropathies. The thyrotoxic effects of cyanide depend on its conversion to the iodineantagonist, thiocyanate.8. Human cassava-eating populations showed ophthalmological and neurologicalsymptoms which are associated with exposure to HCN, though it is likely that othernutritional or metabolic deficiencies affecting the cyanide detoxification mechanism arealso involved (e.g. sulphate and zinc deficiencies).9. Several epidemiological studies in cassava-eating populations, which established anassociation between cyanide exposure and spastic paraparesis, amblyobia ataxia ortropical ataxia neuropathy (TAN) and possibly goitre have also been considered.However, the data are highly confounded by other nutritional and environmental factors.Adequate long-term toxicity studies in animals fed a diet containing HCN or cyanogenicglycosides are also lacking.10. Limited data from the UK show that the average and high (97.5th percentile) dailyintake of HCN from its use in flavours or flavour ingredients were 46 and 214 μg/person,which correspond to approximately 0.8 and 3.6 μg/kg bw/day respectively. Data from aNorwegian dietary survey show that the average and high (97.5th percentile) dailyintake of HCN among consumers amounts to respectively 95 and 372 μg/person or 1.4and 5.4 μg/kg bw/day. Cassava flour is used as a staple food mainly outside Europe; aconsumption of 200 g/person would lead to an estimated intake level of 30 μg HCN/kg

15


bw for a 60 kg adult. In accordance with the JECFA view such an intake would not beassociated with acute toxicity. The highest level of HCN found in retail marzipan paste is20 mg HCN/kg. Assuming on one sitting a person of 60 kg consumes 100 g marzipancontaining such a level, that intake would be equivalent to 2 mg HCN or to 0.03 mg/kgbw.

The Panel concluded that the current exposure to cyanide from flavouring ingredients(97.5th percentile) is unlikely to give rise to acute toxicity. For chronic exposure theoverall data were not considered adequate to establish a numerical no-observed-adverse-effect level (NOAEL) or Tolerable Daily Intake (TDI) in humans. In view of thelack of adequate data on chronic toxicity, the Panel supports the continued application oflimits for the presence of HCN in foods and beverages (73).

2.5.1.1 Absorption, distribution and excretionAbsorption of cyanide occurs across both mucous membranes and intact skin (74). Themost common inorganic cyanides are readily absorbed through the stomach andduodenum. Absorption by the gastrointestinal mucosa is dependent on the pH of the gutand the lipid solubility of the specific cyanide compound (75). Cyanides are rapidlydistributed to all organs and tissues via the blood. The cyanide concentration is higher inred blood cells than in plasma by a factor of two or three, reflecting cyanide’s tendencyto bind to methaemoglobin. Cyanide may also accumulate in body cells by binding tometalloproteins or enzymes such as catalase or cytochrome c oxidase (74). Cyanidelevels in a woman who died 30 minutes after ingesting 2.5 g NaCN were highest (up to3.2 mg% [3.2 mg/100 g]) in the stomach contents, brain and urine (76). Two otherstudies found the concentration of cyanide to range from 0.2 to 2 mg% in the liver,kidneys and brain of poisoned humans (77) (78). It has been reported that small butsignificant levels of cyanide are present in normal, healthy human organs atconcentrations of ≤0.5 mg/kg (as CN–), owing to the breakdown of cyanogenic foods andtobacco smoke by bacterial action and vitamin B12 (79). Cyanide is rapidly detoxified inthe body by conversion to the much less toxic thiocyanate (CNS) ion. (80). The majorpathway for this conversion is via the intramitochondrial enzyme rhodanase, a liverenzyme that catalyses the transfer of sulphur from a donor to cyanide to formthiocyanate (75) (81). Urine has been found to be the major route of excretion ofthiocyanate. Average urinary excretion of thiocyanate normally ranges between 0.85 and14 mg over a 24-hour period (75). Some cyanide is metabolized directly, and carbondioxide is eliminated in expired air. A small amount of HCN is also eliminated in expiredair (81).

2.5.1.2 Toxic EffectsThe cyanides reviewed in this document are highly lethal to humans; lethal oral doses ofcyanide compounds generally range from 50 to 200 mg CN (0.7 to 2.9 mg/kgbodyweight) (82); based on case reports of poisonings, an average fatal dose of cyanidehas been estimated to be 1.52 mg/kg bodyweight (83). In a two-year chronic toxicitystudy in which male and female rats were fed cyanide at doses as high as 7.5 and 10.8mg/kg bodyweight per day, respectively, no clinical or histopathological effects of any

16


kind were observed (84). Cyanide acts through the inhibition of cytochrome c oxidase inthe respiratory electron transport chain of the mitochondria, impairing both oxidativemetabolism and the associated process of oxidative phosphorylation (85) (86). Surveysin African communities where cassava is a staple crop show a strong correlationbetween cassava consumption and endemic goiter and cretinism. Dietary deficiencies,especially low intake of iodine, may contribute to these effects (87) (88).

2.1.5.3 MechanismCyanide has a high complex affinity to certain heavy metals. It blocks Fe(III) (not Fe(II)!)in cellular respiration enzymes of warm-blooded organisms. The respiratory chain isblocked at the stage of cytochrome oxidase Fe3+ and therefore inhibits the activation ofoxygen for further oxidation processes. The complexation is completely reversible (89).

H2-Donator n

NADH

NAD+ onFADH2

FAD onCyt-Fe2+

Cyt-Fe3+

C

CN-

onCyt-Ox-Fe2+

Cyt-Ox-Fe3+

C

CN-

onO2

4-

O2

Figure 2.3: Interaction of cyanide in the respiration chain (simplified). Blockage ofcytochrome oxidase at oxidation stage 3 (89).

2.1.5.4 DetoxificationCyanide is detoxified by rhodanase. This enzyme is present in various different types oftissue in warm-blooded organisms, especially in the liver. It binds cyanide to sulphurunder formation of the less toxic thiocyanate (ger.: Rhodanid): The reaction is reversedby thiocyanate oxidase. In organisms the equilibrium is on the side of thiocyanate.

rhodanase

CN- + S thiocyanate oxidase

CNS-

Figure 2.4.: Cyanide is detoxified by the enzyme rhodanase. The reaction is rreversed bythiocyanate oxidase. In organisms the equilibrium is on the side of thiocyanate.

Velocity of detoxification is high: about 1mg/kg/h from animal experiment; about0.1mg/kg/h from human after intake of sodium nitroprusside for therapeutic purpose.The lethal dose is considered to be 1mg/kg body weight CN-, the minimal lethal dosetherefore is detoxified within 1-10 hours. The limiting detoxification step is themobilization of sulphur in the intermediate metabolism. The capacity of detoxification

17


can be increased remarkable by intake of mobile sulphur, e.g. as sodium thiosulfateNa2S2O3. Hydrogen cyanide and other cyanides do not tend to cumulate.

2.1.5.5 Symptoms of poisoningThe weak acid hydrogen cyanide is dissociated only to 1.6% at pH = 7.4. and thereforediffuses through cell membranes very easily. The blockade of cytochrome oxidase bycyanide is very fast. After inhalation first resorptive symptoms can occur after a fewseconds. After ingestion of inorganic salts, hydrogen cyanide is formed by stomachhydrogen chloride, released and absorbed by the venae portae circulation.

2.1.5.6 Treatment The treatment of cyanide intoxications is aimed on accelerating the body´s owndetoxification and on complexation of cyanide with other heavy metal containingstructures. 1) Sodium-Thiosulfate (Na2S2O3): to supply sulphur for enzymatic formation ofthiocyanate. (10% sol.; 10-20mL every 10 minutes)2) Formation of methemoglobin: Fe(III) in methemoglobin binds to cyanide, thoughwith a lower affinity than cytochrome oxidase. (Inhalation of amyl nitrite).3) Cobalt compounds: Cyanide binds to hexadentical Co-compounds as Co2-EDTA orcobalamin (vitamin B12) (89).

2.1.6 Physical and chemical properties of stone fruit kernels

2.1.6.1 Physical and chemical properties of apricot kernelsThe percentage of the kernel in the pit of apricot varies from 18.8 to 38.0%, calculated as[(kernel)/(pits + kernels)]×100 (1) (90) (91) (92) (93) (94) (95) (96)(97).Some average dimensions of apricot kernels are listed here:length, 14.0–19.17 mm; width, 9.99–10.20 mm; thickness, 3.3–6.27 mm; geometric meandiameter, 9.89–10.31 mm; and mass, 0.47–0.48 g (90) (95) (98) (99). The 100-kernelweight range is 28.7–65.1 g (1)(94) (95)(100). Apricot kernels contain 48-54% oil, 18-23% protein, 21-35% carbohydrates, 0,8-4,7%crude fiber and 6-22% dietary fiber (101) (102).

Figure 2.5 .: Chemical composition of apricot kernels

18


The fatty acid composition of apricot kernel oil is: Oleic acid: 75%; linoleic acid: 17,5%; palmitic acid: 4,5%; stearic acid: 2%; palmitoleicacid: 0,6-0,9%; arachidic acid: <0,4% and small amounts of eicosenoic acid (101) (102).

Figure 2.6.: Fatty acid composition of apricot kernel oil

The reported contents of unsaturated fatty acids are 91.5–91.8% and 7.2–8.3% ofsaturated fatty acids respectively (90), neutral lipids: 95.7–95.2%, glycolipids 1.3–1.8%,and phospholipids 2.0% (90). The kernel oil contains 11.8 mg/100 g campesterol, 9.8mg/100 g stigmasterol, and 177.0 mg/100 g sitosterol (103).The total and oil composition of other stone fruit kernels do not relevantly differ to eachother. Different stone fruit kernels differ though in their tocopherole compositionpattern. Apricot kernel oil e.g. has a relatively high concentration of γ-tocopherole (22,4mg/100g) and a small concentration of α-tocopherole (0,5mg/100g) which makes itdistinguishable from other stone fruit oils (104).Reported protein content of apricot kernel are very variable and range from 14.1 to45.3% (90) (91) (92) (94) (95) (105) (106) (107) (108). A PAGE study found thatapricot kernel proteins contain 84.7% albumin, 7.65% globulin, 1.17% prolamin, and3.54% glutelin. Nonprotein nitrogen is 1.17%, and other proteins are 1.85% (95)Essential amino acids in apricot kernel constituted 32–34% (105) of the total aminoacids. The major essential amino acids (mmol/100 g meal) were arginine (21.7–30.5)and leucine (16.2–21.6), and the predominant nonessential amino acid was glutamicacid (49.9–68.0) (92).In vitro digestibility values of apricot kernel protein by a pepsin-pancreatin enzymesystem are 96.4 ± 1.2 - 99.1 ± 0.3 % (95).Carbohydrate content of apricot kernel was reported variously as 25.5% (w/w) (109),17.3% (94), and 18.1–27.9% (92). The total sugar content was reported as 4.10% inundefatted kernel and 7.76% in defatted kernel (95). The kernels contain thiamine, riboflavin, niacin, vitamin C (90), α-tocopherole, andrelatively high concentration of δ-tocopherole (32.2 mg/100g) (103). The ranges for

19


mineral content of apricot kernel (mg/100 g dry matter) were: Na, 35.2–36.8; K, 473–570; Ca, 1.8–2.4, Mg, 113–290; Fe, 2.14–2.82; Zn, 2.33–3.15 (90) (96) (110).

2.1.6.2 Chemical composition of sour cherry kernel oilPopa et al. (111) reported that sour cherries kernel oil contain high levels of oleic acid(42.9 %), followed by linoleic acid (38.2 %), while the dominant saturated fatty acidswere palmitic (11 %) and stearic acid (6.4%).

Figure 2.7.: Fatty acid composition of sour cherry oil

2.2 Determination of cyanogenic potential in stone fruit kernels and theirproducts.

2.2.1 Norm method: determination of hydrocyanic acid in marzipan,persipan and their prestages by distillation and titration.

The norm method PV-33-HCN 2007-10 for determining hydrocyanic acid in marzipan,persipan and their pre-products is conducted by distillation and subsequent titration(112) (113).

2.2.2 Norm Method: Animal feeding stuffs — Determination of Hydrocyanicacid by HPLC

Cyanogenic glycosides are extracted with an acid solution. After incubation the pH isadjusted to neutral and β-glucosidase is added for enzymatic decomposition at 38°C.Then cyanide is extracted by steam distillation and collected in a potassium hydroxidesolution. In order to obtain a fluorescent complex cyanide is derivatized with taurine andNDA (2,3-Napthalendicarboxyaldehyde). The cyanide complex then is analyzed withHPLC and fluorescence detection.

20


Range of applicationThe norm method ÖNORM EN 16160:2010 is for determination of bound and freecyanide in animal feeding stuffs and animal feeding stuffs raw material of herbal source.The method is validated from 10 mg HCN/kg to 350 mg HCN/kg. If the method is appliedoutside this range it should be validated. A limit of quantification of 2 mg HCN/kg shouldbe achieved usually (114).

2.2.3 Determination of cyanide content by chemical or enzymatic hydrolysisof the cyanogenic glycosides (CG) and quantification of cyanide.

For qualitative or semi-quantitative determinations of cyanide, the easiest and mostreliable method is a simple test with either Feigl-Anger or sodium picrate paper. Thereactions on which these tests are based as well as other methods for detection ofcyanogenic glycosides are described by Hegnauer (15).Sodium picrate paper is prepared by dipping filter paper into an aqueous solution of0.5% picric acid and 5% Na2CO3, then allowing the paper to dry. The paper may bestored dry but should be moistened just before use (16) (115) (116). The detection limitof picrate paper with plant samples is 0.001-0.002% HCN by weight, which correspondsto about 0.01-0.02% prunasin (15) (117).Feigl-Anger paper is made by mixing equal volumes of solutions of tetrabase [4,4'-tetramethyldiaminodiphenyl-methane] and copper ethylacetoacetate (117) (118). Feigl-Anger paper is somewhat more sensitive than picrate paper (119).These qualitative methods are very useful for screening large numbers of samplesbecause they are simple, fairly specific, and require no equipment other than vials andcorks. These qualitative methods also are useful for monitoring separations ofcyanogenic compounds from TLC plates or chromatography paper during isolation andpurification.Methods for the quantification of cyanide in plant tissues involve hydrolysis of thecyanogenic glycosides and release of HCN, which is trapped in a basic solution anddetermined colourimetrically (117) by the method of Lambert et al. (1975) which candetect amounts of HCN of 5 ng and more (120).Methods for the quantification of particular cyanogenic glycosides have been reviewed(121) (122)(123).Emulsine, the β-glycosidase from almonds, has been used for many qualitative andquantitative cyanide determinations. This commercially available enzyme will hydrolyzeseveral cyanogenic compounds (some very slowly), although it is inactive against others(32). This complex mixture of enzymes (β-glucosidase, [3-glucuronidase, sulfalase, andβ-D-mannosidase) is also commercially available. This enzyme is active against a broadspectrum of cyanogenic glycosides. In contrast, most other β-glycosidases haverelatively specific substrate requirements (71) (124) (125) (126) (127)(128).

2.2.4 Determination of amygdalin by chromatographic methods

2.2.4.1 Determination of amygdalin by high pressure liquid chromatography HPLCChandra and Nair (129) reported a method for quantification of benzaldehyde and itsprecursors amygdalin and mandelonitrile in stone fruit kernels. In their work kernelsfrom Michigan-grown tart cherry Prunus cerasus L. cv. Montmorency were frozen at-40°C, lyophilized, crushed and extracted sequentially with hexane and methanol.

21


Amygdalin, mandelonitrile and benzaldehyde were found in the kernel extracts and werequantified by HPLC (C-18, CH3CN:H2O (70:30v/v) isocratic, 1.5mL/min, 210nm). Thesame method could be used to quantify benzaldehyde and precursors in a start productand to measure purity in the end product. (compare 2.5.2.2.4.)Several different HPLC methods for extraction, including methods for inhibition ofepimerisation and quantification of D-amygdalin and neoamygdalin were described byJa-Yong Koo et al. (130), Joo, Woo-Sang et al. (131) and several others.

2.2.4.2 Determination of amygdalin by gas chromatography GC/MSDue to the low vapor pressure of amygdalin it is determined by gas chromatographyindirectly. Kawai et al (132) (133) reported a method for gas chromatographic enzymicdetermination of amygdalin. In this method, amygdalin is hydrolyzed by β-glucosidase.The liberated benzaldehyde is then converted by pentafluorobenzyloxylamine to its o-pentafluorobenzyl oxime which then is determined by gas chromatography using a flameionization or electron capture detector.

2.2.4.3 Determination of amygdalin by micellar electrokinetic chromatography(MEKC)

Chung et al (134) developed a simple, rapid and reproducible method for thedetermination of D-amygdalin and its epimer by using micellar electrokineticchromatography (MEKC). Separation of D-amygdalin was performed in a 20 mM sodiumborate buffer (pH 8.5) containing 300 mM sodium dodecyl sulphate using a bare fused-silica capillary. The eluates were monitored by the absorbance at 210 nm. The appliedelectric field was 278 V/cm, and the time needed for the separation of D-amygdalin didnot exceed 6 min. The calibration curve for D-amygdalin showed excellent linearity inthe concentration range of 5–500 mg/mL. The migration time and the corrected peakarea show relative standard deviations (n=6) of 0.86% and 1.48%, respectively. The limitof detection (S/N=3) for D-amygdalin was 2 mg/mL. Under acidic and neutralconditions, amygdalin exists only as the D-form; however, under basic conditions, itshows both the D- and L-forms with a concentration ratio of 1:1.3 (D-amygdalin/ L-amygdalin). Results of HPLC, UV–Vis spectrophotometry, and mass spectrometryreconfirmed the identification of D-amygdalin and its epimer. The number of theoreticalplates of D-amygdalin is about 100 000 in MEKC, which is significantly higher than~8000 of HPLC. This method has been successfully applied to the determination ofamygdalin epimers in various apricot kernel extracts and pharmaceutical products(134).

2.2.5 Determination of amygdalin by enzyme-immunoassayIn this study New Zealand White rabbits were immunized with an amygdalin-KLH(keyhole limpet hemocyanin) conjugate and produced anti-sera reactive to amygdalin,proving that amygdalin can behave as a hapten in rabbits. Using this polyclonal antibody,a competition enzyme immunoassay for determination of amygdalin concentration inaqueous solutions was developed. This technique is able to effectively detect abnormallyhigh amygdalin content in various seeds and nuts. Enzyme immunoassay can be used to

22


determine the amount of amygdalin in food extracts, which will allow automatedanalysis with high throughput (135).

2.3 Decontamination methodsCaius Plinius Secundus (better known as Pliny the Elder) stated in his 37-volumeencyclopaedia entitled Naturalis Historia, which was completed shortly after his death in79AD, that the Romans were proud of knowing how to remove bitterness from almondkernels (Pliny the Elder, 77;Álvarez, 1798).

2.3.1 Decontamination by waterAfter blanching and peeling, almonds which contain the cyanogenic glycoside amygdalin(bitter almonds) are debittered by leaking with flowing water. The HCN content isreduced by 80% in 24h and the water content of the almond is raised to 38%. Aprolongation of the process reduces the HCN content only minimally (104). Other stonefruit kernels can be decontaminated analogously. Experiments of debittering in batchprocess for less water consumption are described later. (3.2.1.)Decontamination of the cyanide, benzaldehyde and amygdalin containing water used fordebittering the kernels can be done by addition of Fe(II) ions, e.g. as FeSO4 underformation of ferrocyanide:

Fe2+ + 6 CN- [Fe(CN)6]4-

The ferrocyanide can then be precipitated by addition of K+ and Fe3+:

[Fe(CN)6]4- + K+ + Fe3+ K[Fe2(CN)6] ↓

2.3.2 Decontamination with ethanolAlthough ethanol is a much more expensive solvent, decontamination with this organiccompound can be of interest in certain fields of food production: for amarettoproduction stone fruit kernels are leaked in ethanol/water solutions. Presumably thereis a decontamination taking place during the process. Experiments of decontaminationwith ethanol are described later.

2.3.3 Decontamination by fermentation with β-glycosidase (autodecontamination)

By grinding cyanogenic parts of plants the vacuoles get damaged and amygdalin gets incontact to the corresponding enzymes and hydrocyanic acid is formed (16). Thetransient toxin then can be calcinated by exposure to heat.The traditional preparation of cyanogenic plants, such as cassava (Manihot esculenta),includes grinding and exposure to sunlight and air. Experiments described later show that grinding and exposure to heat is not an effectivemethod for decontamination of stone fruit kernels.

23


2.4 Processing options of stone-fruit kernels containing cyanogeniccompounds

2.4.1 OverviewFigure 2.8 shows a brief overview of processing options of stone fruit kernelsexemplified by apricot kernel processing.

Figure 2.8.: Processing options of apricot kernels containing cyanogenic compounds.(136).

24


The work of Gezer et al. (1) was aimed to study the work process and to determine thework capacity, work effectiveness, energy consumption and labour force requirements ofbasic units such as washing, sorting, breaking and separating units in an apricot pitbreaking and separating plant. The study was carried out in an apricot pit processingplant, founded on aclosed area of 300m2 in Malatya, Turkey.

2.4.1 Separation of the apricot kernels from the pitsBecause of the many similar physical properties between kernels and outer parts of theapricot pits, their separation process could not be done effectively for many years andtherefore it was necessary to do it manually.The apricot pits are collected from farmers in sacks and are taken to the sorting units viaconveyers after being washed. In the sorting unit they are classified in four size groupsand then sent to the breaking unit. Broken pits are carried to separating units in sacksand there, kernels are separated from their shells. In order to improve the effectivenessof the separation process, the mixture is dried before entering to separating unit. Thisdrying process can be natural or artificial. Natural drying is made by exposing themixture to the sun for one day. In artificial drying, the mixture spread on the floor isexposed to fuel-heated air for 1 h.There is a semi-automated process cycle in the plant. The washing and sorting unitswork depending on each other. The connection between them is via helix conveyer. Thebreaking and separation units work independently from each other and other units.While transporting between the units, the products are stored in sacks which are madeof polyester material with a capacity of 50 kg.The sorting unit consists of four cylindrical rotating sieves. Its diameter is 0.65 m, itslength is 6 m and its slope is 4%. The sieves have round holes and its first, second, thirdand fourth level holes are 9.5, 10.5, 11.5 and 12.5 mm in diameter, respectively (Fig.2.5.1).

Figure 2.9.: The schematic perspective of sorting unit (1).

Basically, the breaking unit consists of two breaking rollers which are 100 mm indiameter, 500 mm in length. The distance between rollers can be adjusted for each sizeof pit. This unit works approximately at 200 turns/min (Fig. 2.5.2).

25


Figure 2.10.: Apricot pit breaking unit (1).

The separation unit is manufactured by Ender Degirmen Makinaları in Mersin, inspiredfrom an Italian model. It is a gravity table and successfully separates the kernels fromtheir shells through their difference in specific gravity (Kernel density is 1.05 g/cm3, thedensity of pit shell is 1.19 g/cm3). The gravity table is 1350 mm in width, 2800 mm inlength and 1500 mm in height. With the effects of slope, air pressure and horizontal andvertical vibration; the mixture which is dropped onto table with many fine holes isseparated with the lower dense kernels going to upper parts of the table and the moredense shells going to the lower parts of the table. Unbroken pits collected fromunbroken pit sieve are sent to breaking unit for re-breaking. The separated kernels andthe shells are sacked separately.

26


Figure 2.11.: Separating unit (1)

In both the breaking and the separation unit, the ratio of broken kernels was less than1%. The work capacity (t/h), the work effectiveness (%), the energy consumption(kW/h) and the labour force requirements of the basic units in the plant – washing,sorting, breaking and separation – were determined and given in Table 2.5.1. As shownin Table 2.4.1, the mechanization shows the minimal energy consumption and the plantis working with an effectiveness of 99% for separating the kernels from their shells.

Table 2.2.: Basic parameters of washing, sorting, breaking and separating units (1)

27


In all the working units, electrical motors are used. The number of workers in the plantis five. In the plant approximately 5 tonnes of apricot pits are handled each day and 1000kg kernel is obtained.Again in Malatya region, there is a plant with a daily capacity of 40 to 50 tonnes at 2%loss. However, bitter kernels are processed in this plant and washing process is not done.But, it must be considered that all these values may show differences depending on thetype of apricot, whether its kernel is acrid or sweet, the quality of pits and adjustmentsand specifications of units.These results show that the breaking and separating processes of apricot pits can bemade possible and applicable by means of effective and efficient new methods and withlower energy consumption. However, the labour force requirement of the plant can bereduced by an automated continuous process cycle between sorting, breaking, dryingand separating units. In the plant, all machines must be compatible with each other in terms of work capacity.Drying process must be done by a dryer located between breaking and separating units.This can be done by exposing the mixture to the heated air coming from the upperapparatus to the moving platform on which the mixture is placed. For drying, thisprocess will be enough since only the shell of the pits gets wet during the washingprocess.

2.4.2 Kernel processing Due to the high cyanogenic potential, it is not possible to use unprocessed apricotkernels directly in foods. At least one step of detoxification is necessary.

2.4.2.1 Production of edible apricot oilApricot edible essential oil can be either obtained by extraction, hot pressing and coldpressing.

2.4.2.2 Production of natural benzaldehyde from stone fruit kernels

2.4.2.2.1 PrinciplesBenzaldehyde is the simplest and industrially the most important aromatic aldehyde. Itexists in nature, occurring in combined and uncombined forms in many plants. The bestknown source of benzaldehyde is amygdalin. The odour of bitter almonds and otherstone fruit kernels arises from small amounts of free benzaldehyde formed by hydrolysisof amygdalin (137).

Figure 2.12.: Benzaldehyde [100-52-7], structure

28


2.4.2.2.2 Production of natural benzaldehyde

Zhao et al reported 2007 a patented method (138) to obtain natural benzaldehyde fromapricot seeds. The method comprises boiling the apricot seeds for 3 min, peeling, coldpressing at ≤60°C to obtain defatted almond and almond fatty oil, porphyrizing thealmond, screening, adding hydrochloric acid soln. (with pH = 5), hydrolyzing, soaking at40°C for 60 min, steam distillation for 60 min to obtain bitter almond essential oilcontaining cyanide, adding 1 mol/L caustic soda soln., 0.5 mol/L ferrous sulphatesolution, and saturated lime water solution sequentially, allowing to react at 30°C for 10min, and distilling to detoxify for obtaining the detoxified bitter almond essential oil withbenzaldehyde content >90% and cyanide content <0.5 mg/kg.

2.4.2.2.2.1 Recovery of wastewater containing benzaldehyde by simultaneousdistillation-extraction (SDE)

Pollien et al. (139) reported a method for recovery of flavours from wastewater. The by-product from the treatment of apricot kernels contained 178mg/kg naturalbenzaldehyde and non-volatiles (proteins). A recovery of 54% was achieved. (139)

2.4.2.2.2.2 Base-catalysed production from natural cinnamaldehyde It is possible to obtain benzaldehyde from cinnamaldehyde. The product obtained by thismethod declaration as “natural” is not accepted in Europe as it is in the USA (140).

2.4.2.2.2.3 Extraction by microwave Zhang et al (141) reported the microwave processing method of Semen ArmeniacaeAmarum (Prunus armeniaca var. ansu). The experimental results showed thatamygdalase was completely inactivated and the contents of amygdaloside were notreduced at all (141).

2.4.2.2.2.4 Quantification of benzaldehyde in the product.Chandra and Nair (129) reported a method for quantification of benzaldehyde and itsprecursors in stone fruit kernels. In their work kernels from Michigan-grown tart cherryPrunus cerasus L. cv. Montmorency were frozen at -40°, lyophilized, crushed andextracted sequentially with hexane and methanol. Amygdalin, mandelonitrile andbenzaldehyde were found in the kernel extracts and were quantified by HPLC (C-18,CH3CN:H2O (70:30 v/v) isocratic, 1.5mL/min, 210nm) (129). This method can be used toquantify benzaldehyde and precursors in a start product and to measure purity in theend product.

2.4.2.3 Production of persipanPersipan is a cheaper substitute for marzipan produced from other stone fruit kernelsthan sweet almonds. The manufacturing is carried out analogously to the manufacturingof marzipan, only that the amygdalin and hydrocyanic acid has to be removed. This stepis carried out by cold flowing water for 24 hours (104).Experiments are described later where this process can be improved due to less waterconsumption.

29


2.4.2.4 Antioxidant properties of roasted apricot kernel floursDurmaz and Alpaslan (100) evaluated the antioxidant properties of peeled, defatted androasted apricot kernel flours by determining radical scavenging power (RSP), anti-lipidperoxidative activity (ALPA), reducing power (RP), total phenolic content (TPC),assessed by DPPH test, b-carotene bleaching method, iron (III to II) reducing test andFolin method, respectively. Browning degree of the samples were also measured andfound to increase almost linearly with the roasting time. Contrary to browning degree,RSP, RP and TPC did not increase linearly but shows a maximum for 10 min of roasting at150°C. Roasting reduced the ALPA values, thus unroasted sample showed the highestALPA value. RSP, RP and TPC measurements of all samples, were in high correlation (atleast, r = 0.92).

2.4.3 Processing kernels from other stone fruitsKernels from other stone fruits like peach, cherry, sour cherry, nectarine, etc. can beprocessed in similar ways. The results expected are similar products with littledifferences in odour, flavour and colour.

30


3. Materials

3.1 Sample Materials

Sample

Type Details

A Sour cherry kernels Harvested from a sour cherry tree in a private garden in Vienna,Austria. Year: 2009

B Apricot kernels Wachauer Marille from Obstgarten Reisinger, Mitterndorf am Jauerling1 , A-3620 Spitz an der Donau, Austria. Year: 2009

C Apricot kernels Naturgarten Bio-Aprikosenkerne, purchased at BIOWELT, AmNaschmarkt 326-33, Vienna, Austria. Year: 2009

D Apricot kernels Purchased open (without label in a plastic bag) at Am Naschmarkt496, Vienna, Austria. Origin: Usbekistan. Year: 2009

E Apricot kernels,roasted

Wachauer Marille, Wieser Geröstete Marillenkerne, purchased fromMARKUS WIESER GMBH - ÖSTERREICH3610 Weissenkirchen in der Wachau, Wösendorf 13, AustriaYear: 2009

F Bitter almonds Naturgarten Bittermandeln, purchased at BIOWELT, Am Naschmarkt326-331, Vienna AustriaOrigin: Usbekistan. Year: 2009

G Sweet almonds Purchased open (without label in a plastic bag) at Am Naschmarkt613; Origin: ItalyYear: 2009

H Marzipan filling Apricot marzipan filling from "Marillen-Duett" chocolate, containsapricot kernels, purchased at BIOWELT, Am Naschmarkt 326-331,Vienna AustriaYear: 2009

I Chocolate "Marille mit Krokant", contains 4% apricot kernels, purchased atBIOWELT, Am Naschmarkt 326-331, Vienna, Austria. Year: 2009Year: 2009

J Persipan Self made from apricot kernels from Obstgarten Reisinger, Mitterndorfam Jauerling 1 , A-3620 Spitz an der Donau, Austria. Year: 2009

31


K Cherry kernles From Kirschkernkissen (cherry stone cushion), purchased at Bipa,Erdbergerstraße 52-60, Vienna, Austria. Year: 2010

L Flax seed Purchased without label at Am Naschmarkt 496, Vienna, AustriaOrigin: Usbekistan. Year: 2010

MFlax seed shred Purchased at BIOWELT, Am Naschmarkt 326-331, Vienna Austria.

Year:2010

3.2 Laboratory devices

Photometer: Hitachi U9000

Laser diffraction sensor: HELOS (H1809) & RODOS

Coffee mill: Severin KM3872

3.3 Chemicals

Picric Acid: Purchased from Merck (CAS Number: 88-89-1)

Sadiumcarbonate NaCO3: Purchased from Merck (CAS Number: 497-19-8)

Amygdalin from almonds: Purchased from Sigma-Aldrich (CAS Number: 29883-15-6)

β-Glucosidase fom almonds: Purchased from Sigma-Aldrich (CAS Number: 9001-22-3)

Filter paper for test strip: Schleicher und Schull S&S Rundfilter, Whatmann 3MM

Gas chromatography vials (cm)

Phosphate Buffer solution (pH = 5)

32


4. Methods

4.1 Picrate method for determination of total cyanide in prunus kernelsThis simple and accurate method can be applied on many different kinds of samples ifthese are prepared correctly. It can be used for a rough estimation of cyanide by themeans of a simple test strip and a colour chart. More accurate quantification can then beachieved by photometrical measurements of diluted isopurpurate from the test strip(136).

4.1.1 PrincipleA method of Bradbury (142) for determination of total cyanide in cassava and from Egan(143) and Haque (144) in various foods were evaluated and adapted for application oncertain stone fruit species. The method is based on the reaction of hydrocyanic acid andyellow picrate to red isopurpuric acid (145) (146) on a test strip and then measuredeither semi-quantitative by comparision with a colour chart or diluted and measured bymeans of a photometer.

Figure 4.1: Isopurpurate reaction of picric acid (145)

4.1.2 Sample Preparation and preparation of picrate paper test strips

4.1.2.1 Preparation of picrate paper test stripsSodium picrate paper is prepared by dipping filter paper into an aqueous solution of0.5% picric acid and 5% Na2CO3, then allowing the paper to dry. The paper may bestored dry but should be moistened just before use (16) (115) (116). The filter paperchosen for preparation of test strips depends on the expected concentration. For HCNconcentrations of around 40mg/kg or less Schleicher and Schüll.

4.1.2.2 Preparation of homogeneous food samples like marzipan or persipanHomogeneous food samples need no particular preparation. Due to homogeneity of thesample direct measurement is possible. The necessary sample quantity amounts to600mg: 25-100 mg per measurement and it is recommended to do at least sixmeasurements per sample.Preparation of persipan: The persipan measured in this study was prepared asfollowing: 50g where blanched for 5 minutes at 80°C, then peeled and grinded. 15g

33


powdering sugar was added and the mass was formed to a tube and cut into slicesroasted of approximately 2.5cm, then they were roasted at 180° for 20 minutes.

4.1.2.3 Preparation of apricot kernelsFor measuring the total cyanide content in apricot kernels it is necessary to grind thesekernels before measuring. In order to achieve sufficient homogeneity and avoid too largeheat input, an ideal grinding time of 60 seconds has to be maintained. Furthermore it isimportant to grind a sufficient sample quantity (45-50g) so that the mentioned heatinput does not cause a too large rise in temperature. These values refer to grindingprocesses with a coffee mill (model: Severin KM3872). The heat entry caused by themeal process can lead to the escape of prussic acid before the gas-proof sealing of thesample and thus to falsified results. This can be observed by constant mass loss (withinthe range 0,1mg/s) when weighing the sample on analytical scales. Therefore solelyfresh, thermally untreated and dry kernels are to be used. Aging and thermal treatmentsuch as a freezing, blanching or roasting can lead to a change of the kernel structure andthus to a relieved escape of prussic acid. Too much moisture of the sample, probablycaused by prior freezing, leads to clumping of the grains while grinding and thus to astrongly inhomogeneous sample. Ideally the sample is a fine, white powder which isweight-constant immediately after grinding.

4.1.2.4 Preparation of sour cherry kernelsSour cherry kernels seem to be more robust than apricot kernels since the measuredcyanide concentration was not influenced by the sample quantity. A sample quantity ofless than 1g and a grinding time of 120s lead to acceptable results.

4.1.3 Quantification of the total cyanide content in the prepared sampleA weighed amount (25-100mg) of cyanogenic sample is placed in a small gas-proofclosable glass vial (25 mm diam., 50 mm high), followed by 0.5mL of buffer solution and5 drops of ß-glucosidase solution (2mg/mL).A yellow strip of filter paper (previously prepared as described above) it placed into avial which then is immediately tightly stoppered. The vials are placed in an oven at 40°Cover night for at least 20 hours and a colour change from yellow to orange to red can beobserved. The shade of colour can be compared with that of a colour chart (Figure 3.2) to obtainthe approximate amount of cyanide present.

34


Figure 4.2: Colour chart for estimation of total cyanide content.

The paper is then immersed in water for 30 minutes and the absorbance of the solutionis measured at 510 nm against an identically prepared blank sample developed in theabsence of HCN and cyanogenic glycosides.By means of calibration functions (Figure 3.3 and Figure 3.4) of absorbance vs. HCNcontent the total HCN content can be obtained more accurate. The best correlation wasobtained with polynomic interpolation in second degree (136).

35


Figure 4.3: Calibration function 2.5-40mg/kg HCN equivalents

Figure 4.4: Calibration function 40-1056mg/kg HCN equivalents

Although these calibration curves can be used for quantification of total cyanide contentit is recommended to determine new calibration functions for other photometers inother laboratories.

36


4.1.4 Method validationThe Validation was accomplished according to the Guide to Quality in AnalyticalChemistry - An Aid to Accreditation by CITAC and EURACHEM (147).

4.1.4.1 Working rangeThe working range of the method lies between 2.5 and 40 mg/kg HCN equivalents forS&S Rundfilter filter paper and 40-1056mg/kg HCN equivalents for the Whatmann 3MMblotting paper.

4.1.4.2 LinearityAlthough the correlation coefficient is closer to 1 (R2 = 0,9972 for hydrogen cyanidedetermination below 40mg/kg and R2 = 0,9906 above 40mg/kg) linear interpolationleads to quite applicable correlation as well.The calibration data can be interpolated linear for both kinds of test strips:

Figure 4.5: Linear calibration 2.5-40mg/kg HCN equivalents

37


Figure 4.6: Linear calibration 40-1056mg/kg HCN equivalents

4.1.4.3 Limit of detectionA set of blank tests was measured against water. The detection limit is equivalent to themean blank response plus 3 standard deviations [CITAC] and results in 24mg HCNequivalents/kg.

4.1.4.4 Limit of quantificationThe limit of quantification is defined as the lowest point in the calibration function atwhich an effect was observes (Lowest Observed Effect Level, LOEL) and is 2.5mg/kgHCN in the sample when measuring with the more sensitive S&S picrate paper.

38


4.1.4.5 PrecisionRegarding to the standard deviation, the precision for different concentration wasdetermined:

C(HCN) Precision[mg HCN-Eq./L] [%]1056 8,09%845 4,81%634 6,72%422 7,14%211 8,62%155 14,64%103 4,10%52 9,16%41 8,97%

Table 4.1a.: Method precision

Whatmann 3MM paper

C(HCN) Precision[mg HCN-Eq./L] [%]41 2,61%24,6 4,44%16,4 5,69%10 9,95%8,2 8,12%7,5 10,16%5 12,86%3,9 6,34%2,5 13,41%

Table 4.1b.: Method precision

S&S Rundfilter paper

4.1.4.6 Selectivitya) Possible interference factors in the sample solution:

Chemical or physical interaction of amygdalin or HCN with other substances in thesample, such as ketones, aldehydes, fatty acids, etc. cannot be excluded. Therefore a testseries with addition of a certain amount of amygdalin was done: Sample: Sour cherry kernels, time of grinding = 120s, total sample amount: 612,6mg

Test series A: fresh sample, analysis as described

SampleAmount [mg]

Absorbance at 510nm HCN [mg/kg]

A1 55 1,461 1435A2 54,8 1,528 1528A3 42 1,257 1546A4 38,8 1,107 1424A5 63,1 1,524 1323A6 26,6 0,821 1437A7 35,3 1,089 1533

mean: 1461standard deviation: 74,31 5,10%

Table 4.2a.: Test series without addition of amygdalin.

39


Test series B: fresh sample, addition of 100µL amygdalin solution(206mg HCN-Eq./L)

SampleAmount [mg]

Absorbance at 510nm HCN [mg/kg]

B1 35,3 1,176 1689B2 72,5 1,61 1238B3 41,4 1,139 1383B4 51,7 1,378 1414B5 33,5 1,064 1569B6 20,4 0,771 1738

mean: 1505standard deviation: 176,24 11,70%

Table 4.2b.: Test series with addition of amygdalin.

The results are in the expected range. Therefore possible loss due to interaction with thematrix is negligible.

b) Possible interference factors in the test strip:Substances such as creatinine, lead acetate, tin chloride, amino acids and othersubstances, with which picric acid can react coordinatively, can influence the result.Therefore it is important to make sure the test strip is not contaminated with any ofthese substances.

4.2 DecontaminationDifferent methods of decontamination were tested on apricot and sour cherry samplesand compared.

4.2.1 Decontamination using water

4.2.1.1 Continuous flow decontaminationAfter blanching and peeling, almonds which contain the cyanogenic glycoside amygdalin(bitter almonds) are debittered by leaching with flowing water. The HCN content isreduced by 80% in 24h and the water content of the almond is raised to 38%. Aprolongation of the process reduces the HCN content only minimally (104). Due to theclose relationship and similarity between bitter almonds and other amygdalincontaining prunus kernels, the method was tested for application on apricot and sourcherry kernels.

4.2.1.2 Batch decontaminationThe disadvantage of continuous flow decontamination due to the high needs of freshwater is obvious. Furthermore, the processing of the water used for decontamination forthe production of natural benzaldehyde can be done more easily when the concentrationof benzaldehyde is higher. Therefore decontamination by batch process was tested.

4.2.1 Decontamination using ethanolAlthough ethanol is much more expensive than water the decontamination process with ethanolwas tested due to possible application during amaretto production.

40


4.2.3 Decontamination by heat treatmentGrinded and whole kernels samples were put on a glass surface and stored at 80°C and150°C for different time periods. This method did not show a significant detoxificationeffect (see 4.2.2).

41


5. Results

5.1 Preliminary tests and observations

5.1.1 Apricot

Mass ratio:Species: Apricot (Prunus armeniaca) Cultivar: Wachauer MarilleYear: 2009Purchased at Naschmarkt, Vienna

Two samples of each 19 fruits were taken and the mass of whole fruits, moist and driedstones and kernels were weighed. The kernels where dried to constant mass for 12 daysat 30°C.

Sample Mass total[g]

Mass stones,moist [g]

Mass stones,dry [g]

Mass kernels,moist [g]

Mass kernels, dry [g]

A 989.0 54.0 (5.46%) 40,9 (4.14%) 9.88 (1.00%)B 914.2 52.9 (5.79%) 13.97 (1.53%) 9.24 (1.01%)

Table 5.1.: Mass ratios of apricot kernel. Percentages refer to mass of fresh fruit.

Figure 5.1.: Apricot mass ratio

42


Density:The density values of the different parts of the stone are of interest to separate these from eachother by a floating and sinking process. The density of apricot kernels was estimated bymeasuring mass and volume of a sample:

Mass Volume Density Pit 0.5918g 0.50cm3 1.18g/cm3

Kernel 0.4036g 0.39 cm3 1.03 g/cm3

Table 5.2.: Measured mass and volume and calculated density of crushed apricot kernelsample. Values are inaccurate and can only be used as estimation.

Both, pits and kernel sink in distilled water. In a saturated solution of sodium chloride(density ρNaCl(sat.)=1.202g/mL, (148)) , most kernels swim on the surface of the solutionwhereas most parts of the pits remain sinking. This separation can as well be achievedby a sodium chloride solution of 300g/L. To optimize a separation process based on thisprinciple an ideal sodium chloride concentration could be found in further studies.

5.1.2 Sour CherryMass ratio:The mass ratio between kernel and pits of sour cherry stones was measured. 6 samplesof each 10 stones were taken.

Sample A B C D E FTotal mass [g]

2,5681 2,3322,452

52,564

42,384

72,4166

5Kernel mass [g]

0,75910,562

30,626

40,697

70,569

7 0,6379mass ratio kernels to

stone [%]29,56

%24,11

%25,54

%27,21

%23,89

%26,40

%26,12

%

Table 5.3.: Masses and mass ratio of sour cherry stone and kernel. Every sampleconsisted of 10 stones.

Density

Due to the higher density of sour cherry kernels compared to apricot kernels, sourcherry kernels cannot be separated with a saturated sodium chloride solutionsufficiently. The crushed kernels showed better separation by floating and sinkingprocess in a more dense saturated magnesium sulphate solution.

43


5.2 Total HCN content in stone fruit kernels and foods containing these

Some stone fruit kernels and food samples were tested on their total hydrogen cyanide content.

Sample Type Details total hydrogencyanide content

[mg/kg]

A Sour cherry kernels Harvested from a sour cherry tree in aprivate garden in Vienna, AustriaYear: 2009

1521 ± 121 mg/kg

B Apricot kernels Wachauer Marille from Obstgarten Reisinger,Mitterndorf am Jauerling 1 , A-3620 Spitz ander Donau, Austria. Year: 2009

801 ± 78 mg/kg

C Apricot kernels Naturgarten Bio-Aprikosenkerne, purchasedat BIOWELT, Am Naschmarkt 326-33, Vienna,AustriaYear: 2009

1971 ±153 mg/kg

D Apricot kernels Purchased open (without label in a plasticbag) at Am Naschmarkt 496, Vienna, AustriaOrigin: Usbekistan. Year: 2009

2246 ± 79 mg/kg

E Apricot kernels,roasted

Wachauer Marille, Wieser GerösteteMarillenkerne, purchased from MARKUSWIESER GMBH - ÖSTERREICH3610 Weissenkirchen in der Wachau,Wösendorf 13, AustriaYear: 2009

Sample 1:60 ± 6 mg/kg

Sample 2:28 ± 3 mg/kg

F Bitter almonds Naturgarten Bittermandeln, purchased atBIOWELT, Am Naschmarkt 326-331, ViennaAustriaOrigin: UsbekistanYear: 2009

2402 ± 139 mg/kg

G Sweet almonds Purchased without label at Am Naschmarkt613; Origin: ItalyYear: 2009

below limit ofdetection

H Marzipan filling Apricot marzipan filling from "Marillen-Duett" chocolate, contains apricot kernels,purchased at BIOWELT, Am Naschmarkt 326-331, Vienna AustriaYear: 2009


I Chocolate "Marille mit Krokant", contains 4% apricotkernels, purchased at BIOWELT, AmNaschmarkt 326-331, Vienna, Austria. Year:2009Year: 2009


K Cherry kernles From Kirschkernkissen (cherry stonecushion), purchased at Bipa, Erdbergerstraße52-60, Vienna, Austria. Year: 2010

105 ± 8 mg/kg

44


L Flax seed Purchased open (without label in a plasticbag) at Am Naschmarkt 496, Vienna, AustriaOrigin: Usbekistan. Year: 2010

120 ± 35 mg/kg

MFlax seed shred Purchased at BIOWELT, Am Naschmarkt 326-

331, Vienna Austria. Year:2010113 ± 37 mg/kg

Table 5.3.: Measured kernel and food samples

The concentration of total hydrogen cyanide varies between different species and cultivars. In allapricot and sour cherry kernels high concentrations of hydrogen cyanide was measured.Therefore untreated apricot and sour cherry kernels are no appropriate food for humannutrition or animal feed.

A relatively low concentration in the cultivar Wachauer Marille was measured which indicatesthat, after a step of debittering, the kernels of this cultivar might be an interesting option asalternative food source.

Figure 5.2.: Hydrogen cyanide content of various stone fruit kernels [mg/kg]

45


Figure 5.3.: Hydrogen cyanide content of various foods containing stone fruit kernels [mg/kg]

Self made persipan and some commercial available foods such as roasted apricot kernels, flaxseed, flax seed shred contain considerable concentrations of hydrogen cyanide. Also cherrykernels from cherry stone cushion contain hydrogen cyanide. These cushions are intended to beheated up in microwave ovens so they can act as heat accumulator. It would be of interest tostudy the liberation of hydrocyanic acid of these cushions when used in this way.

Other foods like chocolate and marzipan filling using apricot kernels contain no hydrogencyanide.

5.3 Effects of debittering

5.3.1 Effects of debittering by leaching in water

5.3.1.1 Continuous flow debittering

Continuous flow debittering applied on apricot:Approximately 50g apricot kernels were blanched (80°C, 5 min), peeled and put under flowingwater (43mL/s; 5°C). During the process the mass increased to almost 160%. The kernels wereanalysed after 8h and 24h. After 8h the HCN content decreased from 801.32+/- 77.53 mg/kg to23.78 +/- 0.94 mg/kg which accords to a decrease of 97.0 %. After treatment for 24h the HCNcontent decreased to 12.19+/- 9.46 mg/kg which accords to a decrease of 98.4 %.

Continuous flow debittering applied on sour cherry:Approximately 2g sour cherry kernels were blanched (80°C, 5 min), peeled and put underflowing water (43mL/s; 5°C). Within the process the mass increased to 121%. The kernels wereanalysed after 8h and 24h. After treatment for 8h the HCN content decreased from 1520.79 +/- 120.94 mg/kg to 339.25 +/-15.59 mg/kg which accords to a decrease of 77.7 %. After treatment for 24h the HCN contentdecreased to 263.54+/- 19.20 mg/kg which accords to a decrease of 82.7 %.

46


5.3.1.2 Batch debittering

Batch debittering apricot45g apricot kernels were blanched (80°C/5min) and put in a beaker with 200mL tap water. Thebeaker was closed with a watch glass. After 4h the water was decanted (debittering water 1:DB1) and replaced by fresh 200mL (DB2). After further 4 hours the water was removed and thekernels were analysed on HCN content. After 8h the HCN content decreased from 801.32+/-77.53 mg/kg to 29.46 +/- 1.63 mg/kg which accords to a decrease of 96.3 %.The analysis of the water used for debittering yielded these results:

DB1: 24,3 +/- 1,1 mg/kgDB2: 14,7 +/- 0,4 mg/kg

45g of apricot kernels which were debittered in batch for 24h in 200mL water, the HCN contentdecreased to 23.5+/- 1,7mg/kg which accords to 97.1%.

Batch debittering sour cherryApproximately 2g sour cherry kernels were blanched (80°C/5min) and put in a beaker with20mL tap water. The beaker was closed with a watch glass. After 4h the water was decanted andreplaced by fresh 20mL. After further 4 hours the water was removed and the kernels wereanalysed on HCN content. After 8h the HCN content decreased from 1520.79 +/- 120.94 mg/kgto 363.47 +/- 20.01 mg/kg which accords to a decrease of 76.1 %.2g of apricot kernels which were debittered in batch for 24h in 20mL water, the HCN contentdecreased to 305.31+/- 20.25mg/kg which accords to 79.9%.Compared to the debittering by continuous flow, this method shows to be almost as efficientwhile having much lower water need.

Figure 5.4: Effects of different process variations of debittering stone fruit kernels with water.

Precipitation of CN- in the water used for debitteringIn order to process the cyanide containing water used for debittering of stone fruit kernels, it isnecessary to remove the toxic HCN. One simple way to do this is by precipitation as potassiumferrocyanide:

47


6CN- + Fe2+ + 4 K+→ K4[Fe(II)(CN)6] ↓

To 1mL of the DB1 solution 10 drops of 0,05M KOH and 0,05M Fe(II)SO4 were added. Agreenish-yellow precipitation was observed. The analysis of this solution resulted in 8,86 +/-0,57 mg/L HCN. This accords to a decrease of 63.5%. An optimization of this step could beachieved by further studies.

5.3.2 Effects of debittering by leaching in ethanolDebittering of sour cherry kernels in ethanol for four days (batch) supplied significant results.Detoxification was not as efficient as with water, although this data could be considered duringamaretto production.

5.3.3 Effects of debittering by heat treatmentThe attempt of roasting the kernels at temperatures of 80° for seven days and 150°C for 24 hoursdid not lead to any satisfying results. Debittering of stone fruit kernels by heat treatment is not afunctional method.

Figure 5.5: HCN-decrease of sour cherry kernels leaking in ethanol

5.4 Particle size distribution

For optimizing sample preparation particle size distribution was measured of apricot kernelpowder after grinding for 30 and 60 seconds, respectively. The particle size distributions weremeasured by laser diffraction. Device: HELOS (H1809) & RODOS, Settings: R7: 0.5/18.0...3500µm

The results show that to achieve a useful sample a grinding time of 60 seconds in necessary.Grinding the sample for only 30 seconds leads to wide range particle size distribution, whichleads to irreproducible results of HCN content.

48


Apricot, grinding time 30s

0

10

20

30

40

50

60

70

80

90

100

Ver

teilu

ngss

umm

e Q

3 / %

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

Ver

teilu

ngsd

icht

e q3

*

1 5 10 50 100 500 1000Partikelgröße / µm

Figure 5.6.: Particle size distribution of apricot kernel powder, grinded for 30 seconds.

x10 = 38,51 µmx50

= 170,10 µmx90

= 240,43 µm

SMD= 133,48 µmVMD= 165,84 µm

x50 = 201,15 µm x90 = 283,56 µm First completemomentum (areic):SMD= 91,72 µm

First complete momentum (massic):VMD = 171,18 µm

x16 = 50,08 µm x84 = 273,70 µm x99 = 298,36 µm Particle volumespecific surface area: SV = 0,07 m²/cm³

Particle mass specificsurface area:Sm = 250,63 cm²/g

Mode of massspecific particle sizedistribution:Mod = 275,00 µm

Median of mass specificparticle size distribution:Med = 201,15 µm

Table 5.4.: Particle size distribution data of apricot kernel powder, grinded for 30 seconds.

When measuring this sample problems occurred. The high oil content of the sample led toagglomeration and blockage of the disperser.

49


Apricot grinding time 60s

0

10

20

30

40

50

60

70

80

90

100

Ver

teilu

ngss

umm

e Q

3 / %

0.00

0.25

0.50

0.75

1.00

1.25

1.50

1.75

2.00

2.25

2.50

2.75

3.00

Ver

teilu

ngsd

icht

e q3

*

1 5 10 50 100 500 1000Partikelgröße / µm

Figure 5.7.: Particle size distribution of apricot kernel powder, grinded for 60 seconds.

50


x10 = 77,93 µmx50

= 170,10 µmx90

= 240,43 µm

SMD= 133,48 µmVMD= 165,84 µm

x50 = 170,10µm

x90 = 240,43µm

First completemomentum (areic):SMD= 133,48 µm

First complete momentum (massic):VMD = 165,84 µm

x16 = 101,68 µm x84 = 229,41 µm x99 = 289,56 µm Particle volumespecific surface area: SV = 0,04 m²/cm³

Particle mass specificsurface area:Sm = 374,60 cm²/g

Mode of massspecific particle sizedistribution:Mod = 195,00 µm

Median of mass specificparticle size distribution:Med = 170,10 µm

Table 5.5.: Particle size distribution data of apricot kernel powder, grinded for 60 seconds.

51


6. Discussion

6.1 MethodThe picrate paper method for determination of total cyanide is a simple, accurate andinexpensive method, therefore suitable for analysis of stone fruit kernels and productsmade from these, especially of intermediate products during the process of debittering.Approximate estimations of cyanide content can be done in gas-proof vials by the meansof simple test strips, a colour chart, β-glucosidase and buffer solutions without expensivelaboratory equipment. More accurate cyanide quantification can be then be achieved bydeluting the colour from the test strip in water and subsequent photometricalmeasurements. Even higher accuracy can be performed by other methods such as HPLC.There have been no side reactions observed which would alter the results. In case of highcyanide concentrations, a saturation of the testing strip is possible. This can result in ameasured lower concentration. To avoid this, the maximum measureable concentrationof the testing strip needs to be taken into consideration.

6.2 Sustainabilty of stone fruit kernelsThe advantages of utilizing amygdalin containing stone fruit kernels are obvious. Hugeamounts of stone fruit kernels, which are littered or composted every year, can be usedas a source for high quality food: persipan, stone fruit edible oil and natural bitteralmond essential oil. Persipan is a cheaper alternative product to marzipan. Flavor,quality and composition of both products are similar to each other, whereas persipan ismade of stone fruit kernels instead of almonds. A step of debittering is necessary. This ispossible with water and/or precipitation of cyanide with Fe(II) and can be optimized forless water usage and time consumption.

Stone fruit oil contains many unsaturated fatty acids and high concentrations oftocopheroles (104). Additionally the presence of natural benzaldehyde gives the oils adelicate, marzipan-like fragrance and taste. Due to the insolubility of hydrogen cyanidein apolar solvents like edible oil there is no step of debittering necessary and no dangerof intoxication by stone fruit edible oils. After detoxification the press cake remainingafter production of stone fruit kernel edible oil can be used as animal feed.

Natural benzaldehyde, or bitter almond oil is a natural essential oil which can be used asnatural flavor for many products as an alternative to synthetic benzaldehyde which isused for marzipan aroma in many products. The hard pits then can be used for heatproduction, which can be used as energy to fuel the machines used for separating andprocessing the kernels.

52


6.3 Toxicity of stone fruit kernel productsThe main disadvantage of utilization of stone fruit kernels is their toxicity due to the highcontent of amygdalin. Apricot kernels contain substantially less amygdalin than kernelsfrom sour cherry. Yet both products must be debittered before consumption. The mostefficient and less expensive process for debittering is extraction with cold water donetwice. The cyanide ions within the extraction solution then can be made harmless bycomplexation with Fe(II)-ions and then precipitated with Fe(III)-ions and K+ aspotassium ferrocyanide:

Fe2+ + 6 CN- [Fe(CN)→ 6]4-

[Fe(CN)6]4- + Fe3+ + K+ K[Fe→ 2(CN)6] ↓

The analyzed content of cyanide in commercial available stone fruit preserve exceededclearly the legal limit of 50mg/kg. The consumption of such products is strictly notrecommended for health reasons (149).

Currently there is no obligation to label the contense of cyanogen gylcosides likeamygdalin within the EU. Bitter almonds are usually only commercially available insmall units of 50g. The packaging carries a warning on the label, that states „Only forcooking and baking purposes. Keep out of reach of children. Not fit for rawconsumption.“ Currently bitter apricot kernel are sold for raw consumption, whichcarries the same risks as raw bitter almonds. They are not sold in small units, but inunits of 200g instead. Research by the Lebensmittelinstitut Braunschweig in 2006proved a comparable amount of cyanide in the apricot kernels as in bitter almonds. In sixsamples of different vendors, the amount of cyanide was measured between 1949 mg/kgand 2934 mg/kg. Only one oft he samples carried a warning label, and this label onlystated to keep it out of reach of children. The lowest lethal dosis of cyanide for an adultis between 0.5 to 3.5 mg/kg Bodyweight (EFSA 2004). According to a risk assessment bythe british Comittee on Toxicity, the daily intake of cyanide should not exceed 0.02mg/kg bodyweight, while the expert comittee „Aromastoffe“ published a TemporaryMaximum Daily Intake (TDMI) of 0.023 mg/kg bodyweight in 2005. Thus, within theinvestigated samples, a lethal dosis for an adult could not be reached with the amountoft ten apricot kernels, but the maximum acceptable amount of 0.02 mg/kg body weightwas already exceed with one kernel.

Bitter apricot kernels are promoted for alternative cancer treatment and sold via theinternet and specialized stores for organic food, which is not supported by medicalresearch. A common indication on the label of apricot kernels sold this way ist that„apricot kernel are rich in vitamin B17“. The usage of the term vitamin implicates ahealth benefit and no risk at higher consumption levels. Vitamin B17 is the given namefor the cyanogenic glycoside amygdalin. Amygdalin does not have vitamin properties.Thesamples labelled „Aprikosenkerne bitter“ were classified as hazardous to health in

53


accordance with Artikels 14 Abs. 2a der VO(EG) Nr. 178/2002. Even apricot kernelsmarketed as „Sweet Apricot Kernels“ could contain considerable amounts of cyanide;two samples from turkey contained between 200 and 400 mg/kg. The labellingadvertised them as „a real alternative to almonds. Apricot kernels also taste good as asnack or as a addition for muesli.“ Sweet and bitter apricot kernels are only marginallydiffering in the way the look. A selection of the bitter kernels is not possible, the samplesreceived an objection. Fourteen further samples of „Sweet Apricot Kernels“ of turkishorigin contained under 70 mg/kg, which allowed to classify the consumption asunproblematic (150).

The hydrocyanic acid is created during digestion and can result in a severe, acutepoisoning with the symptoms of cramps, emesis and dyspnea. A high dose can result indeath by respiratory paralysis. Those symproms can already arise by the consumption ofonly a few kernels. Because of this, the recommended intake of bitter apricot kernels foradults should not exceed one or two per day, and it might even be advisable to abstaincompletely from eating the kernels and using products containig amygdalin (151).

In my opinion it would be necessary to distribute more information about these foodsand substances to the people and indicate cyanogenic food products as such. Warningsfor maximal daily consumption should be mandatory on the label of the products.Furthermore, it would be possible to protect people which insist, for whatever reason,on high doses of amygdalin by administering substances with cyanide antidotes effectslike sodium thiosulfate Na2S2O3. The capacity of detoxification can be increasedremarkable by intake of mobile sulphur, e.g. as sodium thiosulfate Na2S2O3.

6.4 Conclusion and outlook

In general, the utilization of stone fruit kernels within the EU can be considered asreasonable, as long as there is an awareness of the protential hazards and as long assuitable precautions are realized. The extraction with water in a batch process is aneconomic process for reducing the amount of amygdalin.

Furthermore, it has been shown that the amount of amygdalin in stone fruit kernels hasstrong variations. Those varitions depend strongly on the species and the variety. Thedependence of the amount of amygdalin on producing region, crop year and furtherfactors like fertilization can be researched in continued studies. The results of thesestudies could improve the cultivation of stone fruits with kernels with a low amount ofamygdalin.

Measuring the conectration of hydrocanaic acid with a picrate test strip is a method thatcan be used in further studies ata reasonable price.

54


7. Abbreviations

Abs absorption

ADGH amygdalin diglucosidase

AH amygdalin hydrolase

bw. bodyweight

CG cyanogenic glycosides

Cyt P450 cytochrome P450

diam. diameter

EC enzyme commission number

etc. et cetera

GC gas chromatography

HPLC high pressure liquid chromatography

LOEL lowest observed effect level

MECK micellar electrokinetic chromatography

MS mass spectroscopy

SDE simultaneous distillation-extraction

spp. subspecies

UDP Uridine diphosphate

UDPG uridindiphosphoglucose

vs. versus

55


References1. The study of work process and determination of some working parameters in an apricot pitprocessing plant in Turkey. Gezer I. and S. Dikilitas. s.l. : Journal of Food Engineering, 2002, Vol.53 (2002), pp. 111–114.

2. NAWARO Cascading Pilot. Mackwitz, H. s.l. : Bundesministerium für Verkehr, Innovation undTechnologie, 2006.

3. VERORDNUNG (EG) Nr. 1334/2008 DES EUROPÄISCHEN PARLAMENTS UND DES RATES. 2008.

4. Bitterness in Almonds. Sanchez-Perez, R., Jørgensen, K., Olsen, C.E. , Dicenta, F. and Møller,B.L. 2008, Plant Physiol. , Vol. 146.

5. Toxic Substances in Crop Plants. Davis, R.H. Cambridge, UK : Royal Society of Chemistry, 1991,pp. 202-225. ISBN 0-85186-863-0.

6. Cyanogenic glycosides. Lechtenberg M., Nahrstedt A. [ed.] R. Ikan. Chichester : John Wiley andSons, UK, 1999, In R Ikan, ed,Naturally Occurring Glycosides., pp. 147–191. 0-471-98602-X.

7. Bitterness of kernels of almond x peach hybrids and their parents. McCarty C.D, Leslie J.W.,Frost H.B. s.l. : Proc Am Soc Hortic Sci, 1952, Vol. 59, pp. 254–258.

8. Cyanogenic compounds. Conn, E.E. 1980, Annu Rev Plant Physiol, Vol. 31, pp. 433–451.

9. Pattern of the cyanide-potential in developing fruits. Frehner M., Scalet M., Conn E.E. 1990,Plant Physiol, Vol. 94, pp. 28–34.

10. Biosynthesis of cyanogenic glycosides, cyanolipids and related compounds. Møller B.L., SeiglerD.S. [ed.] BK Singh. New York : Marcel Dekker, 1991.

11. Tissue and subcellular localization of enzymes catabolizing (R)-amygdalin in mature Prunusserotina seeds. Swain E., Li C.P., Poulton J.E. 1992, Physiol Plant , Vol. 100, pp. 291–300.

12. Tissue level compartmentation of (R)-amygdalin and amygdalin hydrolase prevents large-scalecyanogenesis in undamaged Prunus seeds. Poulton J.E., Li C.P. 1994, Plant Physiol, Vol. 104, pp.29–35.

13. Analisis de glucosidos cianogenicos en variedades de almendro: implicaciones en la mejoragenetica. Arrazola, G. s.l. : Universidad de Alicante, Alicante, Spain, 2002, PhD thesis.

14. Relationship between cyanogenic compounds in kernels, leaves, and roots of sweet and bitterkernelled almonds. Dicenta F., Martınez-Gomez P., Grane N., Martı´n ML, Leon A., Canovas J.A.,Berenguer V. 2002, J Agric Food Chem , Vol. 50, pp. 2149–2152.

15. Cyanogene Verbindungen, Chemotaxonomie der Pflanzen, Vol.7. Hegnauer, R. Basel :Birkhäuser Verlag, 1986, Vol. 7.

16. Secondary Plant Metabolism. Seigler, David S. pp. 273-299.

56


17. Characterization of cyanogenic glycosides, cyanolipids, nitroglycosides, organic nitrocompounds and nitrile glucosides from plants. A.M., Seigler D.S. and Brinker. London : AcademicPress, (1993), pp. 51-93.

18. Cyanogenic glycosides; a case study for evolution and application of cytochromes P450. Bak S.,Paquette S.M., Morant M., Rasmussen A.V., Saito S., Bjarnholt N., Zagrobelny M., JørgensenK., Hamann T., Osmani S., et al. 2006, Phytochem Rev, Vol. 5, pp. 309–329.

19. Cyanogenic glycosides. Conn, E.E. 1969, J Agric Food Chem, Vol. 17, pp. 519–526.

20. Cyanogenic compounds as protecting agents for organisms. Nahrstedt, A. 1985, Plant SystEvol , Vol. 150, pp. 35–47.

21. Cyanogenesis in animal-plant interactions. Jones, D.A. s.l. : Ciba Foundation Symposium. JohnWiley & Sons, Chichester, UK, (1988), pp. 151–170.

22. Plant cytochromes P450: tools for pharmacology, plant protection and phytoremediation.Morant M., Bak S., Møller B.L.,Werck-Reichhart D. 2003, Curr Opin Biotechnol, Vol. 14, pp.151–162.

23. Reconstitution of cyanogenesis in barley (Hordeum vulgare L.) and its implications forresistance against the barley powdery mildew fungus. Nielsen K.A., Hrmova M., Nielsen J.N.,Forslund K., Ebert S., Olsen C.E., Fincher G.B., Møller B.L. 2006, Planta , Vol. 223, pp. 1010–1023.

24. The cyanogenic glucoside composition of Zygaena filipendulae (Lepidoptera: Zygaenidae) aseffected by feeding on wild-type and transgenic Lotus populations with variable cyanogenicglucoside profiles. Zagrobelny M., Bak S., Ekstrøm C.T., Olsen C.E., Møller B.L. 2007a, InsectBiochem Mol Biol , Vol. 37, pp. 10–18.

25. Intimate roles for cyanogenic glucosides in the life cycle of Zygaena filipendulae (Lepidoptera,Zygaenidae). Zagrobelny M., Bak S., Olsen C.E., Møller B.L. 2007b, Insect Biochem Mol Biol, Vol.37, pp. 1189–1197.

26. Cyanogenic glucosides and plant-insect interactions. Zagrobelny M., Bak S., Rasmussen A.V.,Jørgensen B., Naumann C.M., Møller B.L. 2004, Phytochemistry , Vol. 65, pp. 293–306.

27. Metabolism of cyanogenic glucosides in Hevea brasiliensis. Lieberei R., Selmar D., Biel B.1985, Plant Syst Evol, Vol. 150, pp. 49–63.

28. Mobilization and utilization of cyanogenic glucosides: the linustatin pathway. Selmar D.,Lieberei R., Biehl B. 1988, Plant Physiol, Vol. 86, pp. 711–716.

29. Cyanogenic lipids: utilizationtion during seedling development of Ungnadia speciosa. SelmarD., Grocholewski S., Seigler D.S. 1990, Plant Physiol, Vol. 93, pp. 631–636.

30. Recent developments in chemistry, distribution and biology of the cyanogenic glycosides.Nahrstedt, A. Oxford : Hostettmann, Lea Clarendon Press, 1987.

31. Cyanogenic glycosides, in Secondary Plant Products. Conn, E.E. 1981, Vol. 7.

32. beta-Glucosidases: Biochemistry and Molecular Biology, ACS Symposium series 533. Conn, E.E.Washington DC : s.n., 1993, pp. 15-26.

57


33. Cyanogenic compounds and angiosperm phylogenie. Saupe, S.G. New York : s.n., 1981,Phytochemistry and Angiosperm Phylogenie, pp. 80-116.

34. Natural cyclopentanoid cyanohydrin glycosides. Part 18. Gynocardin and cyclopentenylglycinein Rawsonia lucida. Andersen, L., Clausen, V., Oketch-Rabah, H.A., Lechtenberg, M., Adersen,A., Nahrstedt, A., and Jaroszewski, J.W. s.l. : Biochemical Systematics and Ecology, Vol. 29(2001), pp. 219-222.

35. Kapillarelektrophoretische Untersuchungen an cyanogenen Glykosiden. Demmer, P. Köln : s.n.,2004.

36. Isolation and structure elucidation of cyanogenic glycosides. Nahrstedt, A. s.l. : Wennesland,Conn, Knowles, Westley, Wissing Academic Press London New York Toronto, 1981.

37. Constraints on effectiveness of cyanogenic glycosides in herbivore defense. Gleadow, R., M. andWoodrow, I. 2002, Journal of Chemical Ecology, Vol. 28, pp. 1301-1313.

38. Cyanogenesis and foodplants In: Phytochemistry and Agriculture – Proceedings of thePhytochemical Society of Europe,. Nahrstedt, A. s.l. : van Beek, Bretler Clarendon Press, Oxford,1993.

39. Why are so many foodplants cyanogenic? Jones, D.A. 1998, Phytochemistry, Vol. 47, pp. 155-162.

40. Bioactivation of cyanide to cyanate in sulfur amino acid deficiency: relevance to neurologicaldisease in humans subsisting on cassava. Tor-Agbidye, J. et al. 1999, Toxilogical Sciences, Vol. 50,pp. 228-235.

41. Biosynthesis of cyclopentyl fatty acids. Cramer, U. and Spencer, F. 1976, Biochim. Biophys.Acta, Vol. 450, pp. 261-265.

42. The biosynthesis of cyanogenic glucosides in seedlings of cassava (Manihot esculenta Crantz).Koch, B., V.S.Nielsen, B.A. Halkier, C.E.Olsen and B.L. Moller. 1992, Arch. Biochem. Biophys.,Vol. 292, pp. 141-150.

43. Cyanogenic glucosides: the biosynthetic pathway and the enzyme system involved. Halkier,B.A., H.V. Scheller, and Moller, B.L. Chichester : John Wiley and Sons. Ltd., 1988, CibaSymposium, Vol. 140, pp. 49-66.

44. The biosynthesis of cyanogenic glucosides in higher plants. Halkier, B.A. and B.L. Moller.1990, J.Biol.Chem., Vol. 265, pp. 21114-21121.

45. Cyanogenic glycosides. Møller B.L., Poulton J.E. s.l. : Academic Press, London, 1993.

46. Cyanogenic and phenylpropanoid glucosides from Prunus grayana. Shimomura, H.Y. Sashida,and T. Adachi. 1987, Phytochemistry, Vol. 26, pp. 2365-2366.

47. Prunasin-6'-malonate, a new cyanogenic glucoside from Merremia dissecta. Nahrstedt, A.,P.S.Jensen and V.Wray. 1989, Phytochemistry, Vol. 28, pp. 623-624.

48. Sur la biogene`se du glucoside cyanoge´ne´tique des feuilles de laurier-cerise (Prunus lauro-cerasus). Mentzer C., Favre-Bonvin J. s.l. : C R Acad Sci Ser III Sci Vie, 1961, Vol. 253, pp. 1072–1074.

58


49. Studies on cyanogenic glucoside of Sorghum vulgare. Akazawa T., Miljanich P., and Conn E.E.1960, Plant Physiol , Vol. 35, pp. 535–538.

50. Isolation of the heme-thiolate enzyme cytochrome P-450TYR, which catalyzes the committedstep in the biosynthesis of the cyanogenic glucoside dhurrin in Sorghum bicolor (L.) Moench.Sibbesen O., Koch B., Halkier B.A., Møller B.L. 1994, Proc Natl Acad Sci USA , Vol. 91, pp. 9740–9744.

51. Cytochrome P-450TYR is a multifunctional heme-thiolate enzyme catalyzing the conversion ofL-tyrosine to p-hydroxyphenylacetaldehyde oxime in the biosynthesis of the cyanogenic glucosidedhurrin in Sorghum bicolor (L.) Moench. Sibbesen O., Koch B., Halkier B.A., Møller B.L. 1995, JBiol Chem , Vol. 270, pp. 3506–3511.

52. Cloning of three A-type cytochromes P450, CYP71E1, CYP98, and CYP99 from Sorghum bicolor(L.) Moench by a PCR approach, and identification by expression in E. coli of CYP71E1 as the oxime-metabolizing cytochrome P450 in the biosynthesis of dhurrin. Bak S., Kahn R.A., Nielsen H.L.,Møller B.L., Halkier B.A. 1998, Plant Mol Biol , Vol. 36, pp. 393–405.

53. Transgenic tobacco and Arabidopsis plants expressing the two multifunctional sorghumcytochrome P450 enzymes, CYP79A1 and CYP71E1, are cyanogenic and accumulate metabolitesderived from intermediates in dhurrin biosynthesis. Bak S., Olsen C.E., Halkier B.A., and MøllerB.L. 2000, Plant Physiol , Vol. 123, pp. 1437–48.

54. Isolation and reconstitution of cytochrome P450ox and in vitro reconstitution of the entirebiosynthetic pathway of the cyanogenic glucoside dhurrin from sorghum. Kahn R.A., Bak S.,Svendsen I., Halkier B.A., Møller B.L. 1997, Plant Physiol , Vol. 115, pp. 1661–1670.

55. The UDP-glucose:p-hydroxymandelonitrile-O-glucosyltransferase that catalyzes the last step insynthesis of the cyanogenic glucoside dhurrin in Sorghum bicolor. Isolation, cloning, heterologousexpression, and substrate specificity. Jones P.R., Møller B.L., Hoj P.B. 1999, J Biol Chem, Vol. 274,pp. 35483–35491.

56. The in vitro substrate regiospecificity of recombinant UGT85B1, the cyanohydringlucosyltransferase from Sorghum bicolor. Hansen K.S., Kristensen C., Tattersall D.B., JonesP.R., Olsen C.E., Bak S.,Møller B.L. 2003, Phytochemistry , Vol. 64, pp. 143–151.

57. Resistance to an herbivore through engineered cyanogenic glucoside synthesis. Tattersall D.B.,Bak S., Jones P.R, Olsen C.E., Nielsen J.K., Hansen M.L., Hoj P.B., Møller B.L. 2001, Science , Vol.293, pp. 1826–1828.

58. Metabolic engineering of dhurrin in transgenic Arabidopsis plants with marginal inadvertenteffects on the metabolome and transcriptome. Kristensen C., Morant M., Olsen C.E., EkstrømC.T., Galbraith D.W., MøllerB.L., Bak S. 2005, Proc Natl Acad Sci USA, Vol. 102, pp. 1779–1784.

59. Immunocytochemical localization of prunasin hydrolase and mandelonitrile lyase in stems andleaves of Prunus serotina. Swain E., Poulton J.E. 1994a, Plant Physiol , Vol. 106, pp. 1285–1291.

60. Enzymatic formation of beta-cyanoalanine from cyanide. Floss H.G., Hadwiger L., Conn E.E.1965, Nature , Vol. 208, pp. 1207–1208.

61. Utilization of amygdalin during seedling development of Prunus serotina. Swain E., PoultonJ.E. 1994b, Plant Physiol , Vol. 106, pp. 437–445.

59


62. The Arabidopsis thaliana isogene NIT4 and its orthologs in tobacco encode beta-cyano-L-alanine hydratase/nitrilase. Piotrowski M., Schönfelder S., Weiler E.W. 2001, J Biol Chem , Vol.276, pp. 2616–2621.

63. Cyanide metabolism in higher plants: cyanoalanine hydratase is a NIT4 homolog. PiotrowskiM., Volmer J.J. 2006, Plant Mol Biol , Vol. 61, pp. 111–122.

64. Evolution of heteromeric nitrilase complexes in Poaceae with new functions in nitrilemetabolism. Jenrich R., Trompetter I., Bak S., Olsen C.E., Møller B.L., Piotrowski M. 2007,Proc Natl Acad Sci USA , Vol. 104, pp. 18848–18853.

65. Maize nitrilases have a dual role in auxin homeostasis and b-cyanoalanine hydrolysis.Kriechbaumer V., Park W.J., Piotrowski M., Meeley R.B., Gierl A., Glawischnig E. 2007, J ExpBot , Vol. 58, pp. 4225–4233.

66. The natural occuring cyanogenic glycosides. Seigler, D. Oxford : s.n., 1977, Progress inPhytochemistry, Vol. 4, pp. 83-120.

67. Recent developments in the chemistry and biology of cyanogenic glycosides and lipids. Seigler,D.S. 1981a, Rev. Latinoamer. Quim., Vol. 12, pp. 39-48.

68. Cyanogenic glycosides and lipids: structural types and distribution. Seigler, D.S. London : s.n.,1981b.

69. Corrected structures of passicoriacin, epicoriacin and epitetraphyllin B and their distribution inthe Flacourtiaceae and Passifloracae. Seigler, D.S. and K.C.Spencer. 1989, Phytochemistry, Vol.28, pp. 931-932.

70. Cyanogenic variation in some Midwestern U.S. plant species. Aikman, K.D., Bergmann,J.E.Ebinger, and D.Seigler. 1996, Biochem. Syst. Ecol., Vol. 24, pp. 637-645.

71. Specifity of action of allelochemicals: Diversification of glycosides in: Allelochemicals: Role inAgriculture and Forestry. Spencer, K. [ed.] G.R. Wallner. Washington D.C. : s.n., 1987. ACSSymposium Series. Vol. 330, pp. 275-288.

72.Localization and catabolism of cyanogenic glycosides, in Cyanide Compounds in Biology . J.E.,Poulton. [ed.] D. Evered and S. Harnett. s.l.: John Wiley & Sons, Ltd., 1988. Ciba Symposium . Vol.140, pp. 67-91.

73. Opinion of the Scientific Panel on Food Additives, Flavourings, Processing Aids and Materials in Contact with Food (AFC) on hydrocyanic acid in flavourings and other food ingredients with flavouring properties. Question number EFSA-Q-2003-145; Adopted on 7 October 2004

74. Diagnosis and treatment of human poisoning. Ellenhorn, M.J. and Barceloux, D.G. New York :Elsevier Science Publishing Company, 1988, Medical toxicology.

75. Toxicological profile for cyanide. Draft for public comment. Prepared for the U.S. Public HealthService by Technical Resources, Inc.,. (ATSDR), Agency for Toxic Substances and DiseaseRegistry. s.l. : Oak Ridge National Laboratory, 1988.

60


76. Areview of cyanide concentrations found in human organs — a survey of literature concerningcyanide metabolism, “normal,” nonfatal, and fatal body cyanide levels. Ansell, M. and Lewis, F.A.S.1970, J. Forensic Med, Vol. 17, pp. 148-155.

77. The content of cyanide in humans from cases of poisoning with cyanide taken by mouth. With acontribution to the toxicology of cyanides. Halstrom, F. and Moller, K.O. 1945, Acta Pharmacol,Vol. 1, p. 18.

78. Cyanide poisoning. Curry, A.S. 1963, Acta Pharmacol. Toxicol.,, Vol. 20, p. 291.

79. The determination of cyanide in biologic fluids by microdiffusion analysis. Feldstein, M. andKlendshoj, N.C. 1954, J. Lab. Clin. Med.,, Vol. 44, p. 166.

80. Hydrogen cyanide and cyanides. Anonymous. 1983, Saf. Pract., Vol. 1, p. 22.

81. Cyanide health advisory. U.S. Enviromental Protection Agency. s.l. : Office of Drinking Water,1985.

82. Health effects criteria document for cyanide. U.S. Enviromental Protection Agency. s.l. :Office of Drinking Water, 1985.

83. Drinking water criteria document for cyanide. U.S. Enviromental Protection Agency. s.l. :Office of Drinking Water, 1987.

84. Chronic toxicity for rats of food treated with hydrogen cyanide. Howard, J.W. and Hanzal, R.F.1955, J. Agric. Food Chem, Vol. 3, p. 325.

85. Cyanide poisoning: an uncommon encounter. Holland, D.J. 1983, J. Emerg. Nurs.,, Vol. 9, p. 138.

86. Handbook of poisoning: prevention, diagnosis and treatment. Dreisenbach, R.H. andRobertson, W.O. Norwalk : Appleton and Lange, 1987.

87. Possible role of cyanide and thiocyanate in the etiology of endemic cretinism. Ermans, A.M.,Delange, F., Van Der Velden, M. and Kinthaert, J. 1972, Adv. Exp. Med. Biol., Vol. 30, p. 455.

88. Role of cassava in the etiology of endemic goiter and cretinism. Ermans, A.M., Mbulamoko,N.M., Delange, F. and Ahlvarartia, R. s.l. : International Development Research Centre, Ottawa,1980, p. 182.

89. Allgemeine und spezielle Pharmakologie und Toxikologie. Forth, W., Henschler, D. and andW., Rummel. s.l. : Bibliographisches Institut & F.A. Brockhaus AG, 1987. ISBN-3-411-03150-6.

90. Apricot Kernel: Physical and Chemical Properties. Alpaslan M. and Hayta, M. Malatya,Turkey : Inönü University Department of Food Engineering, 2006.

91. Chemical Composition of Different Varieties of Apricots and Their Kernels Grown in LadakhaRegion. Kappor, N., K.L. Bedi, and A.K. Bhatia. 1987, J. Food Sci. Technol, Vol. 24, pp. 141–143 .

92. Characterization of the Seed Oil and Meal from Apricot, Cherry, Nectarine, Peach and Plum.Kamel, B.S., and Y. Kakuda. 1992, J. Am. Oil Chem. Soc. , Vol. 69, pp. 493–494 .

93. Chemical and Physical Properties of Apricot Kernel, Apricot Kernel Oil and Almond Kernel Oil.Hallabo, S.A.S., F.A. El-Wakeil, and M.K.S. Morsi. 1977, Egyptian J. Food Sci., Vol. 3, pp. 1–6 .

61


94. Composition of New Zealand Apricot Kernels. Beyer, R., and L.D. Melton. 1990, N.Z. CropHortic. Sci. , Vol. 18, pp. 39–42 .

95. In vitro Digestibility, Physico-chemical and Functional Properties of Apricot Kernel Proteins,.Abd El Aal, M.H., M.A. Hamza, and E.H. Rahma. 1986, Food Chem. , Vol. 19, pp. 197-212 .

96. Composition of Some Apricot (Prunus armeniaca L.) Kernels Grown in Turkey. Özcan, M. 2000,Acta Aliment., Vol. 29, pp. 289–293 .

97. Evalution of Apricot Kernel Proteins. Moussa, M.M., and M.A.B. Taiseer. 1980, Alexandria J.Agric. Res. , Vol. 28, pp. 133–137 .

98. Mechanical Behaviour of Apricot Pit Under Compression Loading. Vursavus, K., and F.Özgüven. J. Food Eng., Vol. 65, pp. 255–261.

99. Some Physical Properties of Hacıhaliloglu Apricot Pit and Its Kernel. Gezer, ˙I., H.Hacıseferoˇgulları, and F. Demir. 2002, J. Food Eng. , Vol. 56, pp. 49–57 .

100. Antioxidant properties of roasted apricot (Prunus armeniaca L.) kernel. Alpaslan, D. et al.2007, Food Chemistry , Vol. 100, pp. 1177–1181.

101. Some characteristics and fatty acids composition of wild apricot (Prunus pseudoarmeniaca L.)kernel oil. Kaya, C., et al. 4, 2008, Asian Journal of Chemistry, Vol. 20, pp. 2597-2602.

102. Chemical composition of bitter apricot kernels from Ladakh, India. Dwivedi, D. H. and Ram,R. B. s.l. : Proceedings of the International Symposium on Plants as Food and Medicine, 2006,2008, Acta Horticulturae, pp. 335-338.

103. Determination of Tocopherols and Sterols by Capillary Gas Chromatography. Slover, H.T., Jr.,H.R. Thompson, and G.V. Merola. 1983, J. Am. Oil Chem. Soc. , Vol. 60, pp. 1524–1528.

104. Lehrbuch der Lebensmittelchemie, 6. Auflage. Belitz, H. D. , W. Grosch, P. Schieberle. s.l. :Springer Verlag, 2008. 3540732012.

105. Chemical Composition of Bitter and Sweet Apricot Kernels. Femenia, A., C. Rossello, A.Mulet, and J. Canellas. 1995, J. Agric. Food Chem. , Vol. 43, pp. 356–361 .

106. Unconventional Protein Sources: Apricot Seed Kernels. Gabrial, G.N., F.I. El-Nahry, M.Z.Awadalla, and S.M. Girgis. 1981, Z. Ernährungswiss., Vol. 20, pp. 208–215.

107. The Chemistry of Prunus armeniaca. Joshi, S., R.K. Srivastava and D.N. Dhar. 1986, Br. FoodJ. , Vol. 88, pp. 74–78, 80 .

108. Egyptian Apricot Kernels Seeds, . Salem, S.A., and F.M.A Salem,. 1973, Dtsch. Lebensm.Rundsch., Vol. 69, pp. 322–324.

109. Toxicological, Nutritional and Microbiological Evaluation of Tempe Fermentation withRhizopus oligosporus of Bitter and Sweet Apricot Seeds. Tunçel, G., M.J.R. Nout, L. Brimer, and D.Göktan. 1990, Int. J. Food Microbiol., Vol. 11, pp. 337–344 .

110. Apricot Stone Kernels as a Valuable Commercial By-product. Normakhmatov, R., and T.Khudaishukurov. 1973, Konservn. i Ovoshchesush. Prom., Vol. 10, pp. 32–33.

62


111. Characterization of sour cherries (Prunus cerasus) kernel oil cultivars from Banat. Popa, V.M.,et al. 2011, Journal of Agroalimentary Processes and Technologies , Vol. 17(4), pp. 398-401.

112. Anlage zur Akkreditierungsurkunde DAP-PL-3942.00 nach DIN EN ISO/IEC 17025:2005.Prüfwesen, DAP Deutsches Akkreditierungssystem. 12247 Berlin : IfP Privates Institut fürProduktqualität GmbH Teltowkanalstraße 2, 2007.

113. DAP. [Online] http://www.dap.de/anl/PL394200.pdf.

114. ÖNORM EN 16160 Ausgabe: 2010-11-15 Futtermittel — Bestimmung von Blausäure mittelsHPLC. Institute, Austrian Standards. Wien: Austrian Standards Institute, 2010.

115. Official of Analysis of the Assoc. of Official Analytical Chemists. Horwitz, W. (ed). WashingtonD.C.: s.n., 1980, p. 433.

116. Poisonous Plants of the U.S. and Canada. Kingsbury, J.M. 1964.

117. Methods for the detection and quantitative determination of cyanid ein plant material.Brinker, A.M. and Seigler, D.S. 1989, Phytochem. Bull., Vol. 21, pp. 24-31.

118. Determination of cyanide and cyanogenic glycosides of plants. Brinker, A.M. and D.S.Seigler. Berlin : Springer Verlag, 1992, Modern methods of plant analysis.

119. Absence of cyanogenesis from Droseraceae. Nahrstedt, A. 1980: s.n., Phytochemistry , Vol.19, pp. 2757-2758.

120. Stable reagents for the determination of cyanide by modiefies König reaction,. Lambert, J.L., J.Ramasamy and J.V.Paukstelis. 1975, Anal. Chem., Vol. 47, pp. 916-918.

121. Determination of cyanide and cyanogenic compounds in biological systems. Brimer, L.Chinchester : John Wiley and Sons, Ltd., 1988, Cyanide, Vol. 140, pp. 177-200.

122. Cyanogenic glycosides and cyanohydrins in plant tissue. Qualitative and quantitativedetermination by enzymatic post column cleavage and electrochemical detection after seperationby HPLC. Brimer, L. and L.Dalgaard. 1984, J.Chromatogr., Vol. 303, pp. 77-88.

123. Methods of liberating and etimating hydrocyanic acid from cyanogenic plant material inCyanide in Biology. Nahrstedt, A., N.Erb and H.Zinsmeister. London: s.n., 1981, Academic Press,pp. 461-471.

124. Isolation and characterization of two cyanogenic ß-glucosidases from flax seeds. Fan, T.W.and E.E. Conn. 1985, Arch. Biochem. Biophys., Vol. 243, pp. 361-373.

125. A rapid and sensitive spectrophotometric assay for prunasin hydrolase activity employingpurified mandelonitrile lyase. Gross, M., G.H. Jacobs and J.E. Poulton. 1982, Anal. Biochem., Vol.119, pp. 25-30.

126. The aglycone specifity of plant ß-glycosidases. Hösel, W. and E.E. Conn. 1982, TrendsBiochem. Sci., Vol. 7, pp. 219-221.

127. Spezifische Glucosidasen für das Cyanglucosid Triglochinin-Reinigung und Charakterisierungvon ß-Glucosidasen aus Alocasia macrorrhiza. Hösel, W. and A. Nahrstedt. 1975, Hoppe-Seyler´sZ. Physiol. Chem., Vol. 356, pp. 1265-1275.

63


128. Characterization of ß-glucosidases with high specifity for the cyanogenic glucoside dhurrin inSorghum bicolor (L.) Moench seedlings. Hösel, W., I. Tober, S.H. Ecklund and E.E. Conn. 1987,Arch. Biochem. Biophys., Vol. 252, pp. 152-162.

129. Quantification of benzaldehyde and its precursors in montmorency cherry (Prunus cerasus L.)kernels. Chandra, A. and Nair, M. G. 1993, Phytochemical Analysis, Vol. 4, pp. 120–123.

130. Quantitative determination of amygdalin epimers from armeniacae semen by liquidchromatography. Koo, J.Y., et al. s.l. : Elsevier B.V.,, 2005, Journal of Chromatography, B:Analytical Technologies in the Biomedical and Life Sciences , Vol. 814(1), pp. 69-73.

131. Prevention of epimerization and quantitative determination of amygdalin in ArmeniacaeSemen with Schizandrae Fructus solution. Joo W.S., Jeong J.S., Kim H., Lee Y.M., Lee J.H., HongS.P. 12 2006, Arch Pharm Res, Vol. 29(12), pp. 1096-101.

132. Improved gas chromatography of amygdalin and its diastereomer. Kawai S., Y. Takayama.1980, Journal of Chromatography A, Vol. 197 (2), pp. 240-5. 0021-9673.

133. Gas chromatographic enzymic determination of amygdalin. Kawai, S., K. Kobayashi, Y.Takayama. 2, s.l. : Elsevier B.V., Journal of Chromatography A, Vol. 210 , pp. 342-5.

134. Micellar electrokinetic chromatography for the analysis of D-amygdalin and its epimer inapricot kernel. Kang, S.H., Jung, H., Kim, N., Shin, D.H. and Chung, D.S. 2000, Journal ofChromatography A, Vol. 866, pp. 253–259.

135. Detection of Abnormally High Amygdalin Content in Food by an Enzyme Immunoassay. A-Yeon Cho, Kye Sook Yi, Jung-Hyo Rhim, Kyu-Il Kim, Jae-Young Park, Eun-Hee Keum, JunhoChung, and Sangsuk Oh. 2006, Molecules and Cells, Vols. 21, No 2, pp. 308-313.

136. Blausäure in Steinobstkernen – Quantifizierung und Dekontamination. Ingrid Steiner,Anatol Desser. Österreichusche Lebensmittelchemikertage 2010 – WertbestimmendeInhaltsstoffe in Lebensmitteln, Tagungsband ISBN 978-3-900554-67-5. pp-13-17

137. Benzaldehyde. Brühne, F. and Wright, E. s.l.: Wiley-VCH Verlag GmbH & Co. KGaA,Weinheim, 2005.

138. Method for producing bitter almond essential oil. Zhao, Z., et al. 2007, Patent200910022665.

139. Simultaneous distillation-extraction: preparative recovery of volatiles under mild conditionsin batch or continuous operations. Pollien, P., et al. Nestle Research Centre, NESTEC Ltd,Lausanne, Switz. : John Wiley & Sons Ltd, 1998, Flavour and Fragrance Journal, Vol. 13(6), pp.413-423.

140. www.roempp.com . [Online] 12 2009.

141. Microwave processing method of semen Armeniacae Amarum. L., Zhang, Q., Yao and H., Xie.1991, China journal of Chinese materia medica , Vol. 16(3), pp. 146-7.

142. Picrate paper kits for determination of total cyanogens in cassava roots and all forms ofcyanogens in cassava products. Bradbury, M.G., Egan, S.V. and J.H. Bradbury. 1999, Journal ofthe Science of Food and Agriculture, Vol. 79, pp. 593-601.

64


143. Simple Picrate Paper Kit for Determination of the Cyanogenic Potential of Cassava Flour.Egan, S.V., Yeoh, H.H. and Bradbury, J.H. Division of Botany and Zoology, Australian NationalUniversity, Canberra, ACT 0200, Australia : s.n., 1998, J Sci Food Agric.

144. Total cyanide determination of plants and foods using the picrate and acid hydrolysismethods. Haque, M.R. and Bradbury, J.H. Division of Botany and Zoology, Australian NationalUniversity, Canberra, ACT 0200, Australia : s.n., 2002, Food Chemistry , Vol. 77, pp. 107–114.

145. Spectrophotometric Determination of Trace Amounts of Free Cyanide in Prussian Blue.Bulcke, G.J. Willekens and A. Van Den. Brussels : s.n., 1979, Analyst, pp. 525-530.

146. Zur Kenntnis der sogenannten Isopurpurat-Reaktion der Pikrinsäure. Auterhoff, H. andHeinzelmann, M. 8, 1971, Archiv der Pharmazie, Vol. 304, pp. 624–627.

147. Guide to Quality in Analytical Chemistry - An Aid to Accreditation. CITAC. s.l. : CITAC andEurachem, 2002.

148. Merck Index, 12th Ed., #8742. 1996.

149: AGES. Presseinformation der AGES vom 3.8.2006. [Online] 29.3.2010.http://ages.at/ages/ueber-uns/presse/presse-archiv/2006/wien-382006-ages-die-marille-ist-zum-essen-da-ihr-kern-nicht/ (bezogen am 29.3.2010)

150: http://www.laves.niedersachsen.de/portal/live.php?navigation_id=20051&article_id=73476&_psmand=23 (bezogen am 24.2.2015)

151: http://www.bfr.bund.de/de/a-z_index/amygdalin-9426.html (bezogen am 24.2.2015)

65

Documents

Diplomarbeit · Ich erkläre weiters Eides statt, dass ich meine Diplomarbeit nach den anerkannten Grundsätzen für wissenschaftliche Abhandlungen selbstständig ausgeführt habe